Premenstrual Syndrome, Premenstrual Dysphoric Disorder, and Beyond: A Clinical Primer for Practitioners

Transcription

1 CLINICAL GYNECOLOGIC SERIES: AN EXPERT S VIEW We have invited select authorities to present background information on challenging clinical problems and practical information on diagnosis and treatment for use by practitioners Premenstrual Syndrome, Premenstrual Dysphoric Disorder, and Beyond: A Clinical Primer for Practitioners Susan R. Johnson, MD, MS The management of adverse premenstrual symptoms has presented a difficult challenge for clinicians. However, based on numerous well-designed research studies over the last decade, we now have diagnostic criteria for the severe form of the syndrome, premenstrual dysphoric disorder, and a variety of evidence-based therapeutic strategies. This review presents a comprehensive, practical description of what the clinician needs to know to diagnose and treat adverse premenstrual symptoms at all levels of severity. Diagnostic criteria are described in detail, including a discussion of the distinction between premenstrual dysphoric disorder and premenstrual syndrome (PMS). The rationale for including prospective symptom calendars as a routine part of the diagnostic evaluation of severe symptoms is presented. The differential diagnosis of cyclic symptoms, including depression and anxiety disorders, menstrual migraine, and mastalgia, and an approach for the management of each of these problems are presented. A treatment approach is recommended that matches the treatment to the degree of problems the woman is experiencing. Serotonin reuptake inhibitors are the treatment of choice for severe symptoms, and most women with PMS/ premenstrual dysphoric disorder will respond to intermittent, luteal phase only therapy. Ovulation suppression should be reserved for women who do not respond to other forms of therapy. The role of oophorectomy is limited, and guidelines for its use are presented. (Obstet Gynecol 2004;104: by The American College of Obstetricians and Gynecologists.) From the Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, Iowa. The authors thank the following individuals who, in addition to members of our Editorial Board, will serve as referees for this series: Dwight P. Cruikshank, MD, Ronald S. Gibbs, MD, Gary D. V. Hankins, MD, Philip B. Mead, MD, Kenneth L. Noller, MD, Catherine Y. Spong, MD, and Edward E. Wallach, MD. The problem of adverse premenstrual symptoms has been recognized in the medical literature for nearly a hundred years, but it is only relatively recently, since 1983, that there has been a consensus regarding operational criteria for a premenstrual syndrome (PMS) and that rigorous scientific methodology has been used to study phenomenology, pathophysiology, and therapy. 1 The results of these 20 years of work can be summarized as follows: Approximately 40% of menstruating women experience luteal phase symptoms that are bothersome; for 25%, these are annoying but do not impair functioning; for 10 15%, the symptoms are severe; and among this latter group, a smaller number (3 5% of menstruating women) report significant impairment of one or more areas of daily life. 2,3 The characteristic symptoms are a mix of mood and physical and cognitive disturbances; the hallmark symptom is irritability. The syndrome is distinct from the depression or anxiety disorders. 4 The sequence of pathophysiological events begins with ovulation, which triggers, by an as yet unknown pathway, a series of changes in neurotransmitters, and among the most important of these is a luteal phase reduction in serotonergic function. 5 Randomized trials confirm that the most effective therapeutic approaches are the use of drugs that either inhibit serotonin reuptake or suppress ovulation. 6 Diagnosis of the severe syndrome is usually straightforward, and proven therapies can be offered. However, many women whose situation does not fit this scenario seek help, including women with less severe symptoms that are bothersome but may not require serotonin reuptake inhibitor therapy, women with either mood or physical symptoms that are adversely affected by the menstrual cycle but who do not have PMS/premenstrual VOL. 104, NO. 4, OCTOBER by The American College of Obstetricians and Gynecologists /04/$30.00 Published by Lippincott Williams & Wilkins. doi: /01.aog e 845

2 dysphoric disorder, and women with severe PMS/premenstrual dysphoric disorder who do not respond to serotonin reuptake inhibitor therapy. Women commonly consult gynecologists for the evaluation of premenstrual symptom problems. For this reason, The American College of Obstetricians and Gynecologists has recently updated its Practice Bulletin on this topic. 7 This article focuses on the evaluation of women with undifferentiated menstrual cycle related complaints and provides information that will assist the practitioner in addressing the full range of menstrual cycle related problems. Evidence-based advice is provided when it is available, and when it is not, the advice is based on over 2 decades of clinical experience and is prefaced with the phrase in my experience to make it clear that the approach is based on anecdotal evidence. NOMENCLATURE OF PREMENSTRUAL SYNDROME AND PREMENSTRUAL DYSPHORIC DISORDER The most common diagnostic labels used for clinically significant luteal phase symptoms have been premenstrual tension and premenstrual syndrome (PMS). In 1987, after a long and difficult debate, the American Psychiatric Association added research diagnostic criteria for severe luteal phase symptoms to the appendix of Diagnostic and Statistical Manual of Mental Disorders, Third Edition Revised (DSM-III-R), labeling this condition late luteal phase dysphoric disorder. In the 1994, these criteria were revised in the fourth edition (DSM-IV-R), and the name was changed to premenstrual dysphoric disorder. 1,8 These criteria require the presence of at least 5 symptoms; and at least one of these must be depressed mood, tension, affective lability, or irritability. Additionally, there must be impairment in at least one area of daily life. Thus, premenstrual dysphoric disorder represents a small subset of women at the extreme end of the PMS severity spectrum. Premenstrual dysphoric disorder has not been shown to be etiologically distinct from less severe adverse premenstrual symptoms. Premenstrual syndrome is mentioned in the DSM- IV-R discussion section for premenstrual dysphoric disorder, but specific diagnostic criteria are not defined, except to say that in premenstrual dysphoric disorder the salient symptoms are the mood problems, whereas in PMS, physical symptoms play a larger role. The implication is that premenstrual dysphoric disorder is primarily a mood problem and PMS primarily a physical one. The empirical evidence does not support this conclusion. The literature instead suggests that adverse premenstrual symptoms generally occur in clusters that include physical, mood, and cognitive symptoms and that the distinguishing feature between groups of women is in the experienced severity and level of impairment, rather than the type of symptom. Thus, in my view, adverse luteal phase symptoms are more accurately classified as mild, moderate, or severe; premenstrual dysphoric disorder is simply severe PMS with impairment, and there is no need in clinical practice (perhaps other than for insurance coding purposes) to make a distinction between severe PMS and premenstrual dysphoric disorder. In this article, the term PMS/premenstrual dysphoric disorder is used when the discussion is relevant to all levels of severity, whereas the term severe PMS/premenstrual dysphoric disorder is used when only the severe end of the spectrum is being discussed. Premenstrual Symptoms or Premenstrual Syndrome? To be clear, symptoms that occur repetitively during the luteal (premenstrual) phase of the cycle are nearly universal among women who ovulate. Some are negative, but others are positive (eg, higher energy levels and increased productivity, increased sexual desire, more vivid dreaming, and increased creativity). Based on several large epidemiological studies, it appears that only about 10% of women report no luteal phase symptoms, and approximately 50% experience one or more symptoms that are mild, last only a few days, and are not bothersome. These women should not be diagnosed as having PMS. DIAGNOSIS OF PREMENSTRUAL SYNDROME/PREMENSTRUAL DYSPHORIC DISORDER Although PMS/premenstrual dysphoric disorder is found among menstruating women of all ages, the majority of women who seek care are over 30 years of age. Usually they describe having had symptoms for several years, so it is not clear whether PMS/premenstrual dysphoric disorder worsens with age or whether older women are simply more likely to ask for help for this problem. Premenstrual syndrome/premenstrual dysphoric disorder resolves after menopause, although there is some evidence that perimenopausal symptoms may be more severe in women who have had severe PMS/ premenstrual dysphoric disorder. To establish a diagnosis of PMS/premenstrual dysphoric disorder, symptoms must be 7 1. characteristic of PMS/premenstrual dysphoric disorder, 2. limited to the luteal phase, 3. causing problems for the woman (and, for premenstrual dysphoric disorder, impairment), and 4. not explained better by some other diagnosis. The symptoms associated with PMS/premenstrual dysphoric disorder fall into the categories of mood (eg, irritability, mood swings, depression, anxiety), physical prob- 846 Johnson PMS and Premenstrual Dysphoric Disorder OBSTETRICS & GYNECOLOGY

3 Fig. 1A B. Prospective symptom chart consistent with premenstrual syndrome (PMS)/premenstrual dysphoric disorder. Typical pattern in a case of severe PMS. Symptoms last the entire luteal phase, but stop by the third day of menses. No symptoms are recorded in the follicular (postmenstrual) phase. Patients were asked to circle the days of period, chart the severity of symptoms using symbols, and leave boxes blank when symptoms were absent. Figure has been redrawn from the original chart for clarity and to comply with journal style. (Adapted from Johnson SR. Clinician s approach to the diagnosis and management of premenstrual syndrome. Clin Obstet Gynecol 1992;35: ) Johnson. PMS and Premenstrual Dysphoric Disorder. Obstet Gynecol lems (eg, bloating, breast tenderness, appetite changes, hot flushes, insomnia, headache, fatigue), and cognitive disturbances (eg, feeing confused, poor concentration), and there can be behavioral consequences (social withdrawal, arguments). Other disorders, such as depression or anxiety, or medical disorders, such as migraine, irritable bowel syndrome, or hypothyroidism, must be differentiated because these may be accompanied by similar symptoms. The timing of symptoms should be confirmed with a prospective symptom and menstrual period record ( charts ) kept by the patient for at least 2 cycles. With 2 months of records, the clinical impression of ovulatory cycles can usually be confirmed and the timing of symptoms relative to menses assessed. For PMS/premenstrual dysphoric disorder, symptoms begin anytime in the 2 weeks before the onset of menstrual flow, continue to the onset of menses, and resolve within a day or two after the onset of bleeding (Fig. 1). In my experience, many clinicians resist the advice to get charted information from patients; I continue to recom- VOL. 104, NO. 4, OCTOBER 2004 Johnson PMS and Premenstrual Dysphoric Disorder 847

4 Fig. 2A B. Pattern consistent with menstrual magnification. Symptoms are present the entire month, but worsen in the late luteal phase, peak during menses, and slowly improve in the postmenstrual phase. Patients were asked to circle the days of period, chart the severity of symptoms using symbols, and leave boxes blank when symptoms were absent. Additional symptoms noted on the chart in April (A): Day 1, facial breakout; Day 3, spotting; Day 4, large clots; Day 6, heavy flow; Day 16, energy to take on tasks; Day 21, worked several jobs; Days 27 28, easily set off; Day 29, spotting. Additional symptoms noted on the chart in May (B): Day 1, lack of patience with children, desire to isolate self from family; Day 5, severe depression, super critical, obsessed with problems, desire to isolate self; Day 11, energy to take on tasks; Day 25, facial breakout. Figure has been redrawn from the original chart for clarity and to comply with journal style. (Adapted from Johnson SR. Clinician s approach to the diagnosis and management of premenstrual syndrome. Clin Obstet Gynecol 1992;35: ) Johnson. PMS and Premenstrual Dysphoric Disorder. Obstet Gynecol mend that charts be obtained. First, prospective information is required by the DSM-IV criteria for premenstrual dysphoric disorder because the literature shows that more than half of the women who present with a complaint of severe PMS are found not to have a pure luteal phase pattern based on prospective charts. 8 Second, the charting period is an excellent time to introduce self-help strategies independently of other therapy. Many women, especially those with more moderate symptoms, will respond to these approaches, and so deferring the decision about pharmacological therapy for 2 months is acceptable to most patients. Third, many women find the information personally helpful because it may help them in anticipating and planning their lives. Finally, even if severe symptoms might warrant immediate drug therapy, obtaining charts during the first 2 months of treatment can provide useful diagnostic information if the treatment is ineffective. Most women have had their symptoms for many years, and when the value of the charts is explained, noncompliance is rare. I use a very simple charting system, as shown in Figures 1, 2, and 3. Women are simply asked to record the 5 most bothersome symptoms and to indicate each evening whether the symp- 848 Johnson PMS and Premenstrual Dysphoric Disorder OBSTETRICS & GYNECOLOGY

5 Fig. 3A B. Pattern consistent with coexisting mood disorder and premenstrual syndrome (PMS)/premenstrual dysphoric disorder. The first symptom listed on this chart is depression, and the remainder are other typical physical/mood symptoms associated with PMS/premenstrual dysphoric disorder. In this case, the patient began a tricyclic antidepressant prescribed by her personal physician at the same time that she began charting. Note that over the 2 cycles the symptom of depression resolves, whereas the other symptoms limited to the luteal phase persist. Patients were asked to circle the days of period, chart the severity of symptoms using symbols, and leave boxes blank when symptoms were absent. Figure has been redrawn from the original chart for clarity and to comply with journal style. (Adapted from Johnson SR. Clinician s approach to the diagnosis and management of premenstrual syndrome. Clin Obstet Gynecol 1992;35: ) Johnson. PMS and Premenstrual Dysphoric Disorder. Obstet Gynecol tom was present, and if so, to indicate its severity. The dates of the menstrual periods are circled. Symptom severity is assessed by patient self-report. Problems with relationships at home and at work, parenting, and work performance can be assessed to determine the extent to which the woman s life is disrupted. Just the fact that a woman seeks treatment implies a significant level of severity. There are no laboratory or physical examination findings associated with PMS/premenstrual dysphoric disorder; the examination and any tests should be aimed at ruling out other possible causes of the symptoms. Specifically, there is no value in obtaining levels of any of the reproductive hormones (estradiol, progesterone, folliclestimulating hormone, luteinizing hormone, testosterone, etc) because these levels do not vary between women with and without PMS/premenstrual dysphoric disorder. The diagnostic process can usually be completed in 2 clinic visits, separated by 2 menstrual cycles, as shown in Box 1 ( Diagnostic Approach: Visits Separated by Two Menstrual Cycles ). At the first visit, the presenting symptoms are evaluated, and a differential diagnosis is devel- VOL. 104, NO. 4, OCTOBER 2004 Johnson PMS and Premenstrual Dysphoric Disorder 849

6 oped. During the next 2 cycles, the woman is asked to keep a daily symptom record (chart), and the physician obtains any indicated referrals or diagnostic test results. At the second visit, all information is reviewed, and a management plan is formulated. BOX 1. DIAGNOSTIC APPROACH: VISITS SEPARATED BY TWO MENSTRUAL CYCLES At the first visit: 1. Assess symptoms, and decide if each is consistent with premenstrual syndrome/premenstrual dysphoric disorder; 2. Based on the history and physical examination, decide if some other diagnosis should also be considered; 3. Consider psychiatric/psychological referral if the mood symptoms are associated with vegetative symptoms, significant suicidal ideation, or frequent inability to function; 4. Obtain laboratory tests or consultations that are indicated to assess other diagnoses under consideration. Between the first and second visit: 1. The woman should record her menstrual periods and daily symptom status on a form you provide; 2. The woman should try recommended lifestyle, nutritional interventions, and over-the-counter drugs if she has not already done so. At the second visit: 1. Assess the symptom pattern recorded on the menstrual calendar. a. Is the menstrual pattern consistent with ovulation? b. Are the problematic symptoms confined to the luteal and early menstrual phases? 2. If the calendars and your overall clinical assessment are consistent with premenstrual syndrome/ premenstrual dysphoric disorder, select a therapeutic approach. Approach to Women Using Oral Contraceptive Pills Women using cyclic combined oral contraceptive pills report premenstrual symptoms at about the same rate as women who are spontaneously menstruating. 3 This is despite the fact that in most cycles, even with low dose pills, ovulation is suppressed. One possible explanation is that premenstrual symptoms are triggered by intermittent exposure to any progestin, whether endogenous or exogenous. A large and confusing literature in this area suggests that approximately half of women with PMS who begin oral contraceptives notice no change in their premenstrual symptoms, and of the remaining women, half feel better and half feel worse. Happily, women using oral contraceptive who have significant PMS symptoms appear to respond to the standard PMS/premenstrual dysphoric disorder therapies, and so my practice is to evaluate oral contraceptive users with PMS symptoms in exactly the same manner as with spontaneously menstruating women. 9 DIFFERENTIAL DIAGNOSIS OF PREMENSTRUAL SYNDROME/PREMENSTRUAL DYSPHORIC DISORDER Dysmenorrhea Dysmenorrhea and PMS/premenstrual dysphoric disorder are different disorders, but women commonly confuse the 2 conditions until they are more familiar with the syndromal definitions of each. Primary dysmenorrhea, characterized by lower midline cramping pain and backache on the first day of menses, occurs most frequently in the second and third decades of life, is etiologically related to dysfunction of prostaglandin, and usually responds well to nonsteroidal prostaglandin synthetase inhibitors. Premenstrual syndrome/premenstrual dysphoric disorder, characterized by a cluster of affective and emotional symptoms beginning up to 2 weeks before menses and improving on the first day of bleeding, is more common in the fourth and fifth decades, is related to central serotonergic dysfunction, and responds to selective serotonin reuptake inhibitor (SSRI) therapy. Women with PMS/premenstrual dysphoric disorder sometimes report luteal phase lower abdominal cramps, and women with dysmenorrhea sometimes have systemic symptoms like headache, fatigue, nausea, and diarrhea, and these may begin a day or two before menses. And, of course, some women have both PMS/premenstrual dysphoric disorder and dysmenorrhea. The clinician simply needs to be aware of all of these possibilities so that the most parsimonious interpretation of the symptoms is made. Depression and Anxiety Disorders Many women who seek treatment in specialty clinics for PMS are found to instead have a psychiatric diagnosis, with depressive and anxiety disorders being the most common. These women generally have one or more mood symptoms characteristic of PMS/premenstrual dysphoric disorder that worsen just before or during menses. Thus, they understandably think they may have PMS. Prospective charting shows that the symptoms are present throughout the cycle, although the severity may indeed worsen in the luteal or menstrual phase. This phenomenon is referred to as either premenstrual exac- 850 Johnson PMS and Premenstrual Dysphoric Disorder OBSTETRICS & GYNECOLOGY

7 erbation or menstrual magnification. For the sake of simplicity, I will use premenstrual exacerbation. The diagnosis of premenstrual exacerbation can be made with reasonable confidence if the woman reports that her symptoms are continuous throughout the month, although more severe around menses. A useful question to get at this issue is to ask: How do you feel in the week after your period is over? In other cases, the picture becomes clear only after examining the prospective symptom record. Figure 2 shows a chart from a woman with premenstrual exacerbation. Women who are being treated for either depression or anxiety can present with a variation of premenstrual exacerbation that can be confusing. In this presentation, the psychiatric disorder is adequately treated except during the late luteal or menstrual phase. The symptom calendar will appear to be consistent with PMS/premenstrual dysphoric disorder because there are no symptoms in the follicular phase; in fact, the symptom pattern is the result of a breakthrough from the current medication regimen. Premenstrual syndrome/premenstrual dysphoric disorder can coexist with any psychiatric disorder. There is no foolproof way to distinguish between premenstrual exacerbation and a mood disorder with coexisting PMS/premenstrual dysphoric disorder (Fig. 3). I use the following diagnostic guidelines in my own practice. If a patient s premenstrual symptoms differ from those of the primary disorder and are characteristic of PMS/premenstrual dysphoric disorder, and if they are limited to the premenstrual phase, I suspect that she probably also has PMS/premenstrual dysphoric disorder. I then determine which treatments for PMS/premenstrual dysphoric disorder she has not tried and recommend that we consider one of these first. Most often, that means adding a SSRI to the current regimen. If serotonergic drugs do not help, I use the same approach as I do for premenstrual exacerbation. If a woman with a pre-existing psychiatric diagnosis reports that her premenstrual symptoms are the same as those of her primary disorder, particularly if the symptoms remain significant during the entire menstrual flow, I suspect that she probably has premenstrual exacerbation. The management of premenstrual exacerbation has not been studied in clinical trials. In my practice, I first assess whether the depression or anxiety is being treated optimally, and this usually requires consultation with the physician managing the mood disorder. As a next step, I may recommend increasing dose of the current antidepressant or anxiolytic during the luteal phase of the cycle because the magnified symptoms will sometimes respond to this approach. For example, a woman with depression and persistent symptoms in the luteal phase, while taking fluoxetine 20 mg daily, may improve if the dose is increased to 40 mg for the 2 weeks before menses. If this approach is not helpful, I recommend a trial of ovulation suppression with continuous high-dose progestin, usually medroxyprogesterone acetate 30 mg daily. In my experience, this is often successful, and after 2 or 3 cycles we switch to intramuscular depomedroxyprogesterone acetate 150 mg every 3 months. If this is not tolerated or is unsuccessful, I recommend a 3-month trial of gonadotropin-releasing hormone (GnRH) agonist therapy. This serves a diagnostic as well as therapeutic function, because women who continue to have symptoms are unlikely to have either premenstrual exacerbation or PMS/premenstrual dysphoric disorder. I use the same guidelines (discussed later) for long-term use of GnRH agonist and for oophorectomy, as I do for PMS/ premenstrual dysphoric disorder. Menstrual-Associated Migraine Migraine headache, a condition experienced by as many as 20% of reproductive-age women, can be adversely affected by the menstrual cycle. True menstrual migraine, in which headaches occur only during the late luteal or menstrual phase, is uncommon. However, female migraineurs commonly report that their headaches increase in frequency or severity at those times, and this is the phenomenon called menstrual-associated migraine. My guidelines for managing both menstrual migraine and menstrual-associated migraine are as follows: 1. Proceed with treatment only if you are confident that the headaches are migraines. Menstrual-associated migraines commonly do not have an aura, but they do have the characteristic unilateral throbbing with photophonia, phonophobia, and nausea and last between 4 and 72 hours. If headaches fit this description, if the neurological examination is normal, and if the headaches have been present for many years especially if the headaches began at menarche and/or there is a family history I feel comfortable in proceeding with treatment. If I am uncertain, for any reason, of the diagnosis, then I first refer the woman to her primary care physician or a neurologist for evaluation. 2. Before making a diagnosis of menstrual-associated migraine, get prospective information about the timing and severity of headaches. For this purpose, the woman simply needs to keep notes of the dates of her headaches and the dates of her menstrual flow for several cycles. Headaches that occur in the 4 days before and the 2 days after the first day of menstrual flow fit within the generally accepted time frame considered consistent with either menstrual migraine or menstrual-associated migraine. 3. Initial management should encompass the standard approaches to migraine headache. A review of this topic VOL. 104, NO. 4, OCTOBER 2004 Johnson PMS and Premenstrual Dysphoric Disorder 851

8 is beyond the scope of this article, but interested readers are referred to the comprehensive practice guidelines developed by the American Academy of Family Physicians and the American College of Physicians American Society of Internal Medicine, in collaboration with the American Headache Society. 10 First steps include use of over-the-counter analgesics and avoidance of triggers. For headaches that do not respond, the next step is the use of abortive therapy, with the most effective and specific approach being the use of the triptan class of drugs. Menstrual-associated migraines have been shown in randomized trials to respond to the triptans for abortive therapy, 11 and to nonsteroidal antiinflammatory drugs (NSAIDs) for prophylaxis (eg, naproxen sodium, 550 mg twice daily, beginning at least 3 days before the expected onset of the headache). Finally, if headaches are frequent, prophylactic therapy is often called for, using a wide variety of agents, including blockers, NSAIDs, and calcium channel blockers. Menstrual-associated migraines are particularly amenable to this approach, because the timing of the headache can be anticipated and the prophylactic drug given just in advance of the expected headache. For example, NSAIDs have been shown to be effective for prophylaxis of migraines when administered for 7 days in advance of the headache. Clinicians who are not comfortable managing headaches at this level should refer the patient to a generalist or neurologist. 4. Hormonal treatment for menstrual-associated migraine may be helpful if standard migraine therapy is inadequate. Estrogen withdrawal, as occurs at the end of the luteal phase of the menstrual cycle, is believed to be a potent trigger for migraine headaches, and the hormone interventions are based on this hypothesis. For spontaneously menstruating women, an estrogen bridge can be used, in which supplemental estrogen (eg, 1 mg oral or 0.1 mg transdermal estradiol) is given approximately 3 days before the expected date of menses and continued for a total of 7 days. 12 For women using oral contraceptives, the supplemental estrogen is given during the entire planned placebo week. So-called long cycle therapy can also be used as well, because it has the advantage of minimizing the number of days that a woman is at risk for headache. Oral contraceptives should be avoided in women over 35 years of age who smoke and in women who experience migraines associated with focal neurological deficits. 13 Oral contraceptives should also be avoided in the prevention of migraine with aura, because they are not as effective with this type of migraine, 14 and there may be an increased risk of stoke with this combination. Another type of hormone intervention is to suppress ovulation, and continuous danazol, tamoxifen, and GnRH agonist given for this purpose have been described. Each of these regimens is costly, has potentially significant adverse effects, and none has been adequately tested in clinical trials. In my own practice, I occasionally use GnRH agonist when the headaches are clearly exacerbated in the late luteal phase and when standard headache therapies and the estrogen bridge approach have failed. 15 Cyclic Mastalgia Breast pain, also referred to as mastalgia, is classified as either continuous or cyclic. Cyclic mastalgia is common and frequently presents as part of the PMS/premenstrual dysphoric disorder symptom complex. In the latter situation, it is generally not severe and does not usually require special therapy. When it is the presenting symptom, mastalgia is typically described as moderate-to-severe bilateral, upper outer-quadrant pain, with tenderness and a feeling of heaviness. The pain can begin anytime in the luteal phase and typically remits with the onset of menses. Up to half of these women experience an adverse effect on sexual functioning, and a smaller number have problems with physical activity, social activity, and work. 16 Bilateral, recurrent luteal phase pain in not a characteristic feature of breast malignancy, but this is what most women with this condition fear. Thus, a careful clinical breast examination should always be done as part of the initial evaluation. Mammography should be reserved for women with focal complaints or a palpable mass and for those who meet standard age and family history associated screening criteria. Nonpharmacological treatment measures are generally effective and include the use of a well-fitting, supportive brassiere (eg, sports bra), reduction in caffeine (although the evidence for effectiveness is modest), weight loss, and perhaps, smoking cessation. Vitamins B and E, and evening primrose oil, although commonly recommended, have not been shown in randomized trials to be effective. In addition, most women with this symptom are concerned that they have or are a high risk for breast cancer and experience improvement when given appropriate reassurance regarding this risk. When these measures fail, several options are available. Danazol ( mg daily), bromocriptine (2.5 mg daily), and tamoxifen (10 mg daily) each provide benefit compared with placebo, but each of these has short-term adverse effects and long-term drawbacks. 17,18 When one of these is necessary, a reasonable approach is to treat with one of these drugs for between 3 and 6 months. Additional short courses of therapy can be used if the pain recurs. Treatments under study that may be equally effective, but with fewer adverse effects, include lisuride 19 and topical nonsteroidal antiinflammatory drugs Johnson PMS and Premenstrual Dysphoric Disorder OBSTETRICS & GYNECOLOGY

9 Other Medical Conditions Other medical conditions can become entrained in the menstrual cycle, such that the symptoms specific to these conditions are manifested exclusively or more commonly in the late luteal or menstrual phase. Among the conditions in this category are seizure disorders (so-called catamenial epilepsy ), genital herpes, diabetes, asthma, Raynauds phenomenon, and angina. 2,21 The initial treatment approach should be to maximize standard therapy for the specific condition. If, however, usual therapies are not sufficient to eliminate late luteal phase symptoms, ovulation suppression, as is described for refractory premenstrual dysphoric disorder, should be considered. HIERARCHICAL APPROACH TO TREATMENT OF PREMENSTRUAL SYNDROME/PREMENSTRUAL DYSPHORIC DISORDER Choice of appropriate intervention should be based on 2 principles. First, because PMS/premenstrual dysphoric disorder is a chronic problem that typically does not resolve until menopause, both cost and adverse effects are important factors in choosing treatment. Second, because women experience different degrees of symptom severity, the severity of the treatment approach should be matched to the severity of the symptoms (see Box 2, Hierarchical Approach to Treatment of Premenstrual Syndrome and Premenstrual Dysphoric Disorder ). Nondrug Interventions These approaches are probably most effective for mildto-moderate PMS symptoms and are often used in conjunction with the mineral or herbal supplements described next. Regular moderate aerobic exercise has been shown in clinical and epidemiological studies to be associated with less severe premenstrual symptoms. Behavioral self-help therapies may also be of benefit, such as use of the relaxation response or cognitive behavioral therapy, and these are described in detail in the reviews of Girman et al 22 and Stevinson and Ernst 23 Parry has shown that interventions that alter circadian rhythms, including partial sleep deprivation and bright light therapy, reduce premenstrual mood symptoms, but the role of these strategies in clinical practice is not yet clear. An increase in complex carbohydrate intake during the luteal phase has been shown in randomized trials to reduce the severity of premenstrual mood symptoms. 27,28 Women who choose this strategy should be given specific examples of this type of carbohydrate (eg, whole grains), and they should be advised to reduce BOX 2. HIERARCHICAL APPROACH TO TREATMENT OF PREMENSTRUAL SYNDROME AND PREMENSTRUAL DYSPHORIC DISORDER* Level 1: Premenstrual syndrome, mild to moderate Lifestyle: Aerobic exercise, nutritional changes (reduce caffeine, salt, alcohol; increase complex carbohydrates) Over-the-counter drugs, choose from among: Calcium (1,000 gm) or magnesium (400 gm) supplements daily Chaste berry fruit Level 1a: Premenstrual syndrome, when physical symptoms are the predominate problem choose, based on symptom type: Spironolactone daily, for breast tenderness and bloating Oral contraceptives (regular or long cycle), or depomedroxyprogesterone acetate, for breast pain, cramps, and other abdominal pain Nonsteroidal anti-inflammatory drugs, in the luteal phase, for most physical symptoms Level 3: Premenstrual syndrome/premenstrual dysphoric disorder, when mood symptoms are the predominate problem: A. Symptom-day only selective serotonin reuptake inhibitor (SSRI) B. Daily SSRI C. If the initial SSRI is ineffective or not tolerated, try at least 2 additional SSRI agents (including venlaflaxine) before abandoning this type of agent D. Buspirone in the luteal phase Level 4: Premenstrual dysphoric disorder not responsive to therapy described in levels 1 3 A. Continuous high-dose progestin (eg, oral medroxyprogesterone acetate mg daily, or depomedroxyprogesterone acetate 150 mg every 3 mo) B. Gonadotropin-releasing hormone (GnRH, usual dose), with add back if continued beyond 6 months C. Bilateral oophorectomy (only if GnRH is demonstrated to be effective in this woman and is the only other option) *Move to the next level if the current approach is ineffective over 2 4 cycles. other caloric intake accordingly so as to not gain weight. The common advice to avoid sugar is a logical correlate of this advice, but this approach has not been studied. The role of caffeine in premenstrual symptoms is unclear. Observational studies have found that women who report the VOL. 104, NO. 4, OCTOBER 2004 Johnson PMS and Premenstrual Dysphoric Disorder 853

10 most severe premenstrual symptoms also report the highest caffeine intake, but it is unclear whether the caffeine is causative, aggravates existing symptoms, or is actually being used as self-treatment for the common symptoms of fatigue and reduced concentration. 29 A trial of reducing or eliminating caffeine for several cycles is reasonable. Opinions about the effect of salt are purely anecdotal and probably derive from the belief that the symptoms of bloating and weight gain are due to salt or at least can be improved by reducing its intake. As with caffeine, it may be reasonable to have women experiment with reducing dietary salt, but they should be guided to continue this practice only if improvement is noted. Supplements Dietary supplements have long been a mainstay of PMS therapy, but only a few have proven to be superior to placebo in carefully performed controlled trials. Three comprehensive reviews of this approach have been recently published, and readers interested in this topic are referred to these for details. 22,23,30 Here, I will comment on supplements for which there is clear evidence of benefit or where a commonly used preparation may have potentially dangerous adverse effect. Vitamin Supplements The role of vitamin B6 (pyridoxine) is unclear. The trials of this supplement were conducted in the 1970s and early 1980s, and the methodologies are generally not consistent with current standards. 31 With this caveat in mind, vitamin B6 may be beneficial for some mild symptoms. Women who choose to use B6 should be aware that chronic use of doses in excess of 200 mg per day have been associated with peripheral neurotoxicity. Mineral Supplements Mineral supplements have each been shown to reduce both emotional and luteal phase symptoms. Most of the data are with calcium. In the largest randomized trial, 497 symptomatic women, many of whom were also taking oral contraceptives, were randomized to either placebo or 1,200 mg of calcium carbonate taken daily. 32 After 3 months, there was an overall 48% reduction in total symptom scores in the active therapy group, compared with a 30% reduction in the placebo group. Magnesium (as magnesium pyrrolidone carboxylic acid, 360 mg, or magnesium oxide, 200 mg) has been studied in smaller trials with similar results. 33,34 Calcium is the preferred choice because of the concomitant bone benefit, but for women who do not tolerate calcium, magnesium is an alternative. Herbal Preparations Ginkgo and kava have been proposed, but neither has been tested in controlled trials, and each has potentially severe adverse effects (kava, hepatotoxicity; ginkgo, drug-drug interaction). St John s wort, whose active ingredient is thought to be hypericum, is a logical therapeutic candidate because of its SSRI-like effects and was found to have beneficial effects in one small open trial. 35 However, it has not yet been evaluated in controlled trials, and women should be aware that St. John s wort induces enzyme systems, such as cytochrome P450, that lead to many drug-drug interactions. Chaste berry fruit (Vitex agnus castus) has been shown in placebo-controlled trials to be superior to placebo in a dose of 20 mg daily. 36 Vitex has few adverse effects and can be found in health food stores. There is a single head-to-head trial comparing Vitex to fluoxetine, and although superior to placebo, it was not as effective as fluoxetine. 37 These results support the clinical impression that dietary supplements are most effective for mild-to-moderate PMS and are generally ineffective for premenstrual dysphoric disorder. Natural Progesterone To those who advocate the use of natural (ie, bioidentical) progesterone, the approach is considered a supplement, in that PMS symptoms are thought to arise as a result of inadequate production of progesterone by the corpus luteum. This hypothesis has not been proven, and natural progesterone was not beneficial compared with placebo in 2 large clinical trials, whether administered as a vaginal suppository 38 or orally as micronized progesterone. 39 Topical progesterone has not been specifically studied for the treatment of PMS, but inference from the studies of other routes of administration (especially intravaginally, in which direct bloodstream absorption also occurs) suggests that it should not be beneficial either. Wyatt el al 6 performed a meta-analysis that included all studies of natural progesterone, as well as synthetic progestins, and concluded there was no benefit to any of these therapies. Combined Oral Contraceptives Study results of the effects of oral contraceptives on PMS in the few randomized trials have been mixed. 40,41 Physical symptoms, such as cramps, breast tenderness, and appetite changes, may be alleviated. A new formulation containing ethinyl estradiol and drospirene, a novel progestin that has spironolactone-like activity, has been studied for this condition in the hope that it might be effective However, in the only randomized trial of this drug for premenstrual dysphoric disorder, although there was a trend toward improvement in all symptoms, 854 Johnson PMS and Premenstrual Dysphoric Disorder OBSTETRICS & GYNECOLOGY

11 a statistically significant effect was seen only for the factor that included appetite changes, acne, and food cravings. 45 Thus, this preparation might be appropriate for women who have mild-to-moderate premenstrual symptoms, especially if birth control is desired as well. Selective Serotonin Reuptake Inhibitor Therapy For women with PMS who do not adequately respond to self-help strategies or supplements after 2 or 3 cycles or who present with severe PMS/premenstrual dysphoric disorder, SSRI therapy should be considered. 5 7 Large, well-designed, randomized, placebo-controlled trials of fluoxetine 46 and sertraline 47 and smaller trials of several other serotonin reuptake inhibitors have shown clear benefit compared with placebo for women diagnosed with premenstrual dysphoric disorder, with at least moderately beneficial response rates in 50 60% of women taking active drug. This is not a generic antidepressant effect, because agents with different mechanisms are not effective. 48 The following paragraphs describe what clinicians need to know to prescribe SSRI therapy for PMS/premenstrual dysphoric disorder. 1. Intermittent therapy is usually as effective as continuous therapy. The effect of the SSRIs even fluoxetine with its long half-life on PMS/premenstrual dysphoric disorder is rapid, with symptom improvement usually seen in hours. This differs from the treatment of depression or anxiety, where 4 6 weeks of therapy is required for maximum effect, and suggests that a separate mechanism of action for these drugs must exist. This rapid response suggests that women could be treated with intermittent, rather than continuous therapy, and this has been demonstrated in several randomized trials Intermittent therapy has clear advantages, including lower cost and fewer days at risk for adverse effects (including sexual ones). There are 2 intermittent therapy regimens: luteal phase and symptom day. Luteal phase therapy is started 2 weeks before the expected menstrual period, and symptom day therapy starts on the first day of typical symptoms. In both methods, the drug is discontinued either on the first day of menses or 1 3 days later, depending on the usual day that symptoms spontaneously resolve. If intermittent therapy is inadequate, daily (continuous) SSRI therapy should then be tried. 2. Low-dose therapy is usually adequate. In the large trial of continuous fluoxetine, a dose of 20 mg worked as well as 60 mg, and the rate of adverse effects was lower with the former. 46 The optimal dose of sertraline is not as clear, because of the design of the large-scale trial, but the dose range was between 50 and 150 mg, with a similar level of response as fluoxetine. In my experience, women who require doses at the higher end of the drug s dose spectrum are more likely to have an underlying mood disorder as opposed to a pure PMS/premenstrual dysphoric disorder. 3. If one SSRI is ineffective, do not give up. Because there is no evidence to suggest that one SSRI is superior to another for the treatment of PMS/premenstrual dysphoric disorder, the initial choice can be based on cost, patient desires, and the woman s response and adverse-effect experience. Lack of response to one SSRI may not predict response to other agents in this class, 53 and so my practice is to try at least 3 different drugs before moving to a different form of therapy. Venlaflaxine, which inhibits the reuptake of both serotonin and norepinephrine, should be considered as well. It has been shown to be effective compared with placebo for premenstrual dysphoric disorder, and because of its dual action, it may be beneficial for some women who do not respond to or tolerate the pure SSRIs Physical symptoms improve, not just mood. Women using SSRI therapy also report significantly lower physical symptom scores as well as those for mood. 55 Thus, I typically do not give additional specific therapy for physical symptoms until we assess the overall effect of SSRI therapy over 2 or more cycles. 5. Selective serotonin reuptake inhibitors are usually well tolerated, but be aware of the potential adverse effects. Based on the placebo-controlled, randomized trials of SSRI therapy for premenstrual dysphoric disorder, the most common minor adverse effects include gastrointestinal disturbance (including nausea), insomnia, fatigue, headache, dry mouth, dizziness, tremor, and sweating. These may resolve after several weeks and are not as much of a problem with an intermittent dosing schedule. Longterm continuous use can be associated with sexual side effects, including decreased libido and delayed orgasm. The level of risk is not clear; premenstrual dysphoric disorder itself can have an adverse effect on sexuality, and some women improve after starting SSRI treatment. Intermittent use minimizes this risk. If a woman requires continuous therapy for adequate results, there are not proven strategies for reducing sexual side effects, other than reducing the dose. Some authors have reported benefit from adding a small dose of bupropion (eg, 150 mg daily) to the regimen, and this appears to be safe. Withdrawal symptoms, including dizziness, lethargy, nausea, irritability, lowered mood, and vivid dreams, can occur when long-term SSRI therapy is stopped, but in my experience these are rare with intermittent therapy. A rare serious condition, serotonin syndrome, has been reported in patients receiving an VOL. 104, NO. 4, OCTOBER 2004 Johnson PMS and Premenstrual Dysphoric Disorder 855

12 SSRI along with a monoamine oxidase inhibitor, multiple SSRI drugs, or rarely, with very high doses of a single SSRI. This syndrome has not been reported in any women using low-dose SSRI therapy for PMS/ premenstrual dysphoric disorder. Anxiolytics Anxiolytics are effective for some women when SSRIs are not. Alprazolam and buspirone, given in the luteal phase only, have been found to be superior to placebo. 39,56 Buspirone, compared with alprazolam, has less addictive potential, does not need to be tapered at the end of the luteal phase, has overall fewer adverse effects than alprazolam, and is my preference if I use an anxiolytic. My usual starting regimen is buspirone 5 mg twice daily during the luteal phase, stopping on the first day of menses. If not effective, I increase the daily dose by 10 mg (divided) in subsequent cycles up to no more than the maximum daily dose of 60 mg. Women should be reminded that these agents are central nervous system depressants and that alcohol should generally be avoided while taking them. When Physical Symptoms Are the Primary Problem or Persist After Serotonin Reuptake Inhibitor Therapy The majority of women who seek treatment for luteal phase symptoms have mood symptoms (irritability, mood lability, depressive symptoms). However, a few women will present with a physical symptom as their primary complaint, or more commonly, will continue to have problems with a physical symptom after successful therapy with another approach. Spironolactone ( mg daily) may offer benefit for breast tenderness, bloating, luteal phase weight gain, and possibly, mood symptoms. 57,58 Oral contraceptives have been shown to reduce the physical symptoms associated with PMS, but they do not reliably improve mood symptoms. 59 The nonsteroidal anti-inflammatory agents, given in the luteal phase (eg, naproxen sodium 600 mg 3 times daily), have been shown in several small randomized trials to improve all physical symptoms except breast tenderness, and they may also modify mood symptoms. 59 Follow-up and Duration of Drug Therapy My practice is to see a woman after 2 cycles of a new prescribed therapy, and if it is effective, to see her again 4 months later. If still effective, I revert to annual follow-up. Therapy will likely be needed until menopause, but a trial period off the drug can be considered. In the only study of long-term use of SSRI therapy for PMS/premenstrual dysphoric disorder, most women experienced recurrence when discontinuing the drug after only 1 year. 60 In my own experience, women on long-term therapy often stop the drug for a cycle or two on their own as a test; most recur, but some do not. Thus, I tell women that therapy until menopause is the norm, but that a trial discontinuation sometime after 2 years is reasonable if they like. Ovulation Suppression Patients who do not respond to any of the above regimens are candidates for a trial of ovulation suppression. The gold standard for this approach is GnRH agonist. 61 The problems with long-term use of this agent are high cost, the need for add back estrogen therapy to prevent or minimize bone loss, and the lack of safety data for use beyond a year. Danazol and high-dose continuous estrogen also effectively suppress ovulation and premenstrual dysphoric disorder symptoms, but each has significant adverse effects that make long-term use inadvisable (for danazol, weight gain, androgenic side effects, and profound suppression of high-density lipoprotein; and for estrogen, high risk of endometrial hyperplasia despite progestin). 62 The alternative approach is a suppressive dose of continuously administered progestin. I generally start with medroxyprogesterone acetate, administered continuously as either mg orally every day, or as a depot injection (in the typical contraceptive dose of 150 mg intramuscularly every 3 months). Although this regimen has not been studied in clinical trials, in my view it is worth trying as the initial approach because of its low cost, overall safety, and relative lack of adverse effects. I generally begin with 2 months of oral therapy, and if this is successful, switch to the depo form. If medroxyprogesterone acetate is ineffective, I recommend a trial of GnRH therapy for a minimum of 3 6 months. For the initial trial, I consider add-back hormones as optional and give them primarily for vasomotor symptoms. Regimens that include a cyclic progestin may result in a recurrence of premenstrual symptoms 63,64 ; continuous combined regimens are thus preferable. 15 This has not been a significant problem in my own experience, but when it occurs, I have used progestin-only therapy (eg, daily norethindrone) with good results. If GnRH agonist therapy is effective, a decision must then be made about longer-term therapy. The patient must be informed that this is not a therapy approved by the U.S. Food and Drug Administration and that treatment for longer than a year is not approved for any benign condition. The main known consequence of longer-term therapy is bone loss, and appropriate measures must be taken to minimize this problem, including adequate intake of calcium (1,200 mg daily) and vitamin D (between 400 and 800 mg daily), a baseline bone 856 Johnson PMS and Premenstrual Dysphoric Disorder OBSTETRICS & GYNECOLOGY