8.1.4 Premenstrual syndrome

The premenstrual disorders, premenstrual syndrome (PMS) and premenstrual dysphoric disorder (PMDD) are psychoneuroendocrine disorders characterized by a constellation of affective, somatic, and behavioural symptoms that occur monthly, during the luteal phase of the menstrual cycle with relief soon after the onset of menses. PMS affects approximately 15–40% of reproductive aged women depending on criteria for diagnosis. PMDD is a severe form of PMS, with an emphasis on the affective symptoms. It has been estimated that only 5–8% of women meet the strict criteria for PMDD, but up to 20% may be one symptom short of meeting the criteria (1). The premenstrual syndromes adversely impact relationships, activities of daily living, and workplace productivity.

The research and treatment of the premenstrual disorders have been hampered by lack of consensus regarding the specific diagnostic criteria, methods of assessment of symptoms and impairment, and absence of animal models or biological markers for the disorders. However, elucidation of various aspects of the pathophysiology, well designed multicentre treatment trials, and patient and clinician education have successfully improved diagnosis and management This chapter will review symptoms, definitions, diagnostic criteria, aetiology, evaluation, and nonpharmacological and pharmacological management of PMS and PMDD.

A wide range of symptoms have been attributed to PMS, encompassing emotional, physical, cognitive, and behavioural domains (Box 8.1.4.1). Complicating the picture is the fact that many medical and psychiatric disorders are exacerbated premenstrually or occur as comorbid disorders with PMS/PMDD. Up to 90% of reproductive age women describe some premenstrual symptoms; most are viewed as normal and not troublesome. These mild symptoms are pathognomic of ovulation and are termed moliminal symptoms. There is no consensus as to whether there are core symptoms that define the clinical syndrome of PMS, but the most common complaints accounting for much of the impairment are irritability, tension/anger, anxiety, mood swings, depression, feeling out of control, fatigue, and difficulty concentrating. Irritability and tension may be the cardinal symptoms of PMS. What distinguishes normal moliminal symptoms from those of PMS/PMDD is not the nature, but the severity of these symptoms and their impairment of daily functioning. Similarly, the symptoms are not distinctive in character from those of other medical or psychiatric disorders, but are unique in their cyclic premenstrual timing. Symptoms of PMS/PMDD can be present at any time after ovulation, persist throughout the luteal phase, but generally peak 2–6 days premenstrually and remit within the first 4–5 days of menses. There must be a relatively symptom free phase before ovulation.

The syndrome now known as PMS has been recognized since antiquity. A loose constellation of symptoms was first termed ‘premenstrual tension’ by an American gynaecologist, Frank in 1931. Current diagnostic criteria lack worldwide acceptance of any one particular set of criteria, number of symptoms, duration of symptoms or specific rating scales to operationalize the criteria and degree of impairment (2). The diagnosis of PMS was facilitated by the publication of a set of criteria based on the type, timing and severity/impact of symptoms outlined by the American College of Obstetricians and Gynecologists (ACOG) (3) (Box 8.1.4.2). ACOG diagnostic criteria require at least one of six affective and one of four somatic symptoms is present for 5 or more days before menses for at least three consecutive menstrual cycles. Symptoms must be relieved within 5 days of onset of menses, not occur before the peri-ovulatory phase of the menstrual cycle, present in absence of pharmacological treatment, associated with impairment or dysfunction in social or economic performance, and prospectively confirmed during two menstrual cycles. Other causes of the symptoms must be excluded. These criteria acknowledge the importance of dysfunction and impairment symptoms, and require prospective confirmation of retrospective reporting. Prospective daily rating increases the burden of formulating a diagnosis, but recognizes the potential bias of retrospective recall, a phenomenon termed ‘menstrual magnification’ or the attribution of any symptom that is exacerbated premenstrually to PMS.

The WHO International Classification of Diseases (ICD-10) criteria for PMS are more widely known outside of the United States than are the ACOG criteria. PMS is denoted as premenstrual tension syndrome (PMTS), listed in the section of gynaecological disorders of the female genital organs (4). The ICD-10 diagnosis focuses on the cyclicity and premenstrual timing of the symptoms in association with the menstrual cycle, does not specify a level of severity or impairment, and lacks exclusion criteria; only one of the symptoms from Box 8.1.4.1 must be experienced premenstrually for an ICD-10 diagnosis.The diagnostic criteria for PMDD are delineated in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSMIV) by the American Psychiatric Association (5) (Box 8.1.4.2). PMDD criteria require recurrent cyclic symptoms be present during luteal phase and absent during mid- to late follicular phase for the majority of cycles during the previous year. Five or more of 11 listed symptoms must be present; at least one symptom should be a major mood symptom of moderately severe to severe degree (depression, anxiety/tension, irritability, or affective liability). Symptoms should cause subjective impairment and interfere with work, social activities, and/or relationships, and must not be an exacerbation of another underlying condition, although they may be superimposed on other disorders. Similar to ACOG PMS criteria, symptoms must be prospectively confirmed by daily ratings during at least two consecutive cycles. Physical symptoms are of lesser importance in the DSM criteria than ACOG or ICD 10, and are clustered in one of the 11 items.

The prevalence of PMS and PMDD vary depending on criteria used and method of symptom confirmation. Studies of PMDD suggest 3–8% of menstruating women meet strict DSM criteria for PMDD and 15–20% of menstruating women meet criteria for subthreshold PMDD or severe PMS (1). The prevalence of mild to moderate PMS is less clear and in some studies ranges to 60%. Symptoms often begin in adolescence, peaking in the late 20s to early 30s, persisting until the climacteric. The disorder generally recurs with most ovulatory cycles, spontaneous remission is rare, and symptoms generally relapse within a few cycles after discontinuation of effective treatment.

The morbidity of the disorders is multifactorial, resulting from symptom severity, duration, the chronic repetitive nature, high rate of relapse, and functional impairment in relationships, social activities, and work productivity (6). Women with PMDD endure an estimated 3.8 years of disability over their reproductive years and based on the global burden of disease model, the disability is similar in magnitude to other major medical and psychiatric disorders (7). Impaired functioning also significantly affects the group of women with subthreshold PMDD and severe PMS, expanding the burden of illness to approximately 1 in 5 women of reproductive age (7). Substantial impairment in schoolwork productivity and social life has also been documented in students with PMS/PMDD. Premenstrual symptoms significantly affect health-related quality of life; result in increased health care utilization, work absenteeism, and decreased occupational productivity. Women with PMS were nine times more likely to report one full week of impairment per month causing decreased productivity, and interference with hobbies and relationships, two or more workdays missed for health reasons, significantly increased frequency of ambulatory health care visits, and were more likely to accrue over $500 in health care costs over 2 years (8). Similar to the economic burden with other chronic medical disorders, the economic impact of PMS related to work productivity loss and absenteeism exceeds the direct medical costs (9).

There are few well described risk factors for PMS. Twin studies suggest a genetic predisposition (10).Community samples demonstrated that a history of traumatic events and pre-existing anxiety or panic disorder, current or past history, or family history of major depressive disorders including postpartum depression are more prevalent in women with PMS/PMDD (11). Unipolar major depressive disorders are the most common psychiatric comorbidity among women affected by either PMS or PMDD. Both disorders occur frequently in women, however, research suggests shared biological vulnerability. Generalized anxiety and panic disorders also frequently co-occur with premenstrual disorders. There is a high incidence of panic response in women with PMS given the provocations such as carbon dioxide inhalation or sodium lactate that induce panic symptoms and physiological responses characteristic of panic disorder. Morbidity is further increased when women with psychiatric illness also have superimposed PMS or PMDD. Lifestyle risk factors for PMDD include earlier age of menarche, higher body mass index, cigarette smoking, greater alcohol or caffeine consumption, and higher perceived stress level (12).

The symptoms of PMS and PMDD are triggered in predisposed individuals, by ovulation and subsequent rise and fall of ovarian sex steroids in the luteal phase of the menstrual cycle. Frequent sampling of blood at different phases in the menstrual cycle of women with PMS and controls showed no significant differences in circulating reproductive hormones or peptides including oestradiol, progesterone, testosterone, prolactin, thyroid or adrenal hormones, follicle-stimulating hormone (FSH), luteinizing hormone, thyroid stimulating hormone (TSH), or adrenocorticotropic hormone (ACTH) (13–15). Those women with premenstrual disorders may constitute a subset of women with abnormal central nervous system (CNS) sensitivities to normal fluctuations in hormones that occur during the menstrual cycle. Evidence of altered response to ovarian steroids was shown in a study in which women with PMS and asymptomatic controls were subjected to ‘medical oophorectomy’ utilizing a gonadotropin hormone-releasing hormone (GnRH) agonist, leuprolide. PMS symptoms scores were significantly lower after the leuprolide, but with hormone replacement using physiological doses of either transdermal 17 β oestradiol (0.1 mg) or vaginal progesterone (200 mg twice per day), subjects with PMS reported adverse mood symptoms such as sadness, anxiety, and irritability, whereas the normal women remained asymptomatic (16).

Despite the predominance of mood symptoms and the co-occurrence with affective disorders, PMS does not appear to be a subset of major depressive disorder as evidenced by the differential response to antidepressants and absence of biological markers characteristic of depression (17). Women with PMS experience a rapid improvement in symptoms with serotonergic antidepressants, generally within days, whereas those with depressive or anxiety disorders require 3–4 weeks to achieve a response. PMS sufferers unlike those with affective disorder, respond only to antidepressants that enhance synaptic serotonin, and fail to improve with those antidepressants that augment norepinepherine or dopamine. In contrast to those with depressive disorders, PMS subjects do not demonstrate a failure of dexamethasone to suppress morning cortisol, do not manifest decreased platelet monoamine oxidase B, and do not show abnormal rapid eye movement (REM) latency during sleep electroencephalogram (EEG) studies. Current theories of the pathophysiology of PMS with the most supportive evidence include perturbations of the serotonergic (5-HT) and gamma amino butyric acid (GABA) systems.

Serotonergic dysfunction has been implicated in the aetiology of PMS. Serotonin (5-HT) is a neurotransmitter that also has hormone like effects when released into the bloodstream, regulating mood, behaviour, sexual functioning, appetite, smooth muscle contraction, and the immune systems. A depletion of the serotonergic system causes irritability, poor impulse control, depressed mood, anxiety/ panic, obsessions, compulsions, and cravings for food (18). Clinical evidence for the role of 5-HT in premenstrual disorders include:

PMS does not occur during anovulatory cycles, and appears to be triggered by ovulation, implicating progesterone, the major hormone produced by the corpus luteum. Although there is no evidence for a deficiency of progesterone, a metabolite of progesterone, allopregnanolone has been studied extensively (21). Progesterone is metabolized in the ovary, adrenal gland and brain to form neuro-active steroids, including a key steroid, allopregnanolone (ALLO) (Fig. 8.1.4.1). These 3α, 5α reduced progesterone metabolites can also be synthesized de novo in the brain from the precursor cholesterol, and are referred to as neuroactive steroids because of their effect on brain functioning. ALLO demonstrates potent sedative, hypnotic and anxiolytic properties in rodents and humans and in rodent models of anxiety-like and depressive-like behaviour (18, 22, 23) ALLO and pregnanolone are positive allosteric modulators of the GABA_A receptor, facilitating opening of chloride channels and promoting GABAergic transmission. GABA is the major inhibitory neurotransmitter system in the CNS. Decreased serum ALLO has been found in the luteal phase in women with PMS in some but not all studies. Altered GABA synthesis and activity in the brain have been also been documented in PMS using various imaging techniques (19).

GABA postsynaptic receptors are classified into three subtypes: GABA_A, GABA_B, and GABA_C receptors. The type A receptor is a basic inhibitory control mechanism fundamental to the functioning of the CNS, and regulation of mood and behaviour. The GABA_A receptor is the site of action for the endogenous steroid ALLO, as well as for benzodiazapenes, barbiturates, alcohol, and anticonvulsants (21–24).

In a rodent model simulating some aspects of PMS, progesterone exposure employed to replicate the luteal phase of the menstrual cycle results in anxiety-like behaviour (24). Progesterone administration results in increased production of ALLO in the CNS, and it is the reduced progesterone metabolite ALLO, not progesterone that was determined to be responsible for these behavioural changes. ALLO exposure effects short term changes in GABA_A receptor composition, consisting of up regulation of α4, δ, and γ-subunit expression. This alteration in GABA_A receptor configuration temporarily decreases the sensitivity to GABA and GABA agonists, rendering the GABA_A receptor insensitive to further modulation by neurosteroids. The augmented production of GABA_A receptors with the α4 subunit is associated with the anxiety-like reaction in rodent behaviour al tests. Expression of the γ2 subunit is similarly altered after fluctuations in progesterone. CNS concentration of these neurosteroids, duration of exposure, as well as genetic predisposition are important determinants of the neurosteroid/GABA_A receptor interaction (18, 24).

The alterations of GABA_A receptor isoforms that give rise to reduced neurosteroid sensitivity and CNS excitability and are hypothesized to be important in the aetiology PMS (24). Plasticity of GABA_A receptor population results from fluctuation in endogenous or administered steroids, such as progesterone or ALLO. ALLO concentrations and GABA_A receptor functioning and modulation vary throughout the menstrual cycle, and is likely contribute to the negative mood symptoms experienced by many women during the luteal phase. PMS sufferers have been shown to be insensitive to modulation by benzodiazepines in the luteal phase (24) Women with PMS, but not controls demonstrated decreased responsiveness in the luteal phase to pregnanolone and to the benzodiazepine, midazolam in a novel study using saccadic eye velocity as a proxy for central GABAergic tone (21, 24).

Various lines of evidence support a link between the serotonergic and GABAergic neurotransmitter systems. A salutary interaction between serotonin and GABA in women with PMS was demonstrated in a study by Backstrom et al. in which the lowered luteal sensitivity to midazolam in PMS was increased with the administration of an SSRI to levels observed in asymptomatic women (18, 24). ALLO is also decreased in the cerebral spinal fluid (CSF) of individuals with major depression but increases after SSRI treatment, likely due to stimulation one of the major enzymes involved in the formation of ALLO from cholesterol or progesterone (18). SSRI treatment increases GABA concentration in the occipital cortex and 5-HT receptor agonists enhances GABA_A stimulated chloride ion influx. Serotonergic neurons terminate on inhibitory GABAergic interneurons in the hippocampus and 5-HT receptors have been localized to GABA interneurons in the cortex and hippocampus. In the 5-HT receptor knock-out mice, there is an alteration of GABA_A receptor subunit configuration, GABA_A receptor binding is reduced, and the mice develop benzodiazepine-resistant anxiety. It has also been hypothesized that SSRI treatment can augment GABA-mediated inhibitory processes in the limbic system that are involved emotional regulation (18).

Oestrogen and progesterone appear to be capable of triggering mood disturbances in a susceptible population, and in general seem to modulate PMS symptoms. The involvement of sex steroids is supported by the therapeutic response of women with PMS to suppression of ovulation with gonadotropin releasing hormone (GnRH) agonists (25). Combined oral contraceptives suppress ovulation, the hormonal event responsible for the neurotransmitter alterations, and should be effective for PMS but most studies have not demonstrated efficacy. Studies of women who used different oral contraceptives found little change in premenstrual symptoms, and in over 15% there was a worsening of mood, although approximately 12% did report improvement. Many women on oral contraceptives have bothersome physical and emotional symptoms during the 3 weeks of hormone exposure and others have symptoms during the week off, just before menstruation. The mechanism underlying these symptoms are still unclear, but exposure to and withdrawal from sex steroids and fluid retention mediated by ethinyl oestradiol effects on the renin angiotensin aldosterone system have been implicated. Progestins in the oral contraceptive can adversely modulate GABA_A receptor plasticity (26).

Sex steroids also modify 5-HT availability at the neuronal synapse, affecting serotonergic activity in the periphery and brain (18). The means by which alterations in gonadal steroids trigger changes in the behavioural state of certain individuals are unclear and likely multifactorial. In contrast to pathological function of other endocrine systems (e.g. adrenal, thyroid), where abnormal hormone levels are associated with mood and behavioural disorders, gonadal steroids in the context of normal ovarian function alter central and peripheral neurotransmitter, neuropeptide, and other systems, and precipitate affective and somatic disturbances limited to the luteal phase of the menstrual cycle. Brain imaging studies are likely to be important in the future in understanding changes in brain activity across the menstrual cycle in women with PMS. Functional magnetic resonance studies, while performing cognitive tasks during the follicular and luteal phase in women with and without PMS showed differential responses in the dorsilateral prefrontal cortex and orbitofrontal cortex that have been hypothesized to contribute to altered luteal phase processing of negative and positive stimuli in PMS (19).

There are no specific diagnostic tests for PMS or PMDD. The diagnostic process is one of exclusion based on a clinical interview, physical examination, and prospective daily rating. A comprehensive history should include assessment of medical, gynaecological, and psychiatric, and psychosocial history, life stressors, substance abuse, and domestic violence. The differential diagnosis of PMS is outlined in Box 8.1.4.3. Symptoms such as a headache, fatigue, pelvic or abdominal discomfort and bloating, oedema presenting premenstrually, may be a result of other medical conditions, most commonly chronic pelvic pain, endometriosis, irritable bowel syndrome, migraine, chronic fatigue syndrome, anemia, fibromyalgia, or hypothyroidism. Fibrocystic breast disease, galactorrhoea, and breast cancer should be considered in a woman who presents with predominantly breast symptoms. Symptoms of depressive and anxiety disorders or substance abuse will not demonstrate a purely cyclic pattern, with a consistent asymptomatic phase before ovulation, although these disorders often worsen premenstrually. Past history of depressive or panic episodes, post partum depression, suicidal ideation, substance abuse or failure to confirm luteal phase timing of mood symptoms should prompt a psychiatric consultation.

If a patient is using any hormonal contraceptives, the symptoms experienced prior to the monthly bleeding are not technically PMS, but could be attributed to hormone withdrawal associated with the hormone-free interval or to adverse effects of the hormonal constituents of the contraceptive. Over the age of 40, symptoms such as dysphoria, breast tenderness, headache, and sleep disturbances could be an early manifestation of the climacteric, and should be differentiated from those of PMS by lack of premenstrual timing and elevated day 3 FSH.

A thorough physical examination should be completed including a breast and pelvic examination, but laboratory assessment is not necessary, unless indicated by the history and physical examination. For example, a patient with fatigue, swelling, and weight gain associated with moodiness should prompt TSH screening and possibly complete blood count to rule out anemia and blood chemistry for electrolytes, renal and liver function.

The patient should then be educated on the importance of daily prospective charting of relevant symptoms for a period of 2 months. One of the validated rating scales such as the Daily Record of Severity of Problems (Fig. 8.1.4.2), visual analogue scales (VAS), or the Calendar of Premenstrual Experiences (COPE). An improvised rating scale can also be created consisting of daily scoring of a list of the patient’s five worst symptoms, from 0 to 3 (none, mild, moderate, severe), circling the days of menstrual bleeding, and adding an asterisk for days with impairment. Prospective charting of symptoms has multiple benefits: (1) facilitates the diagnosis of affective or anxiety disorders that might require psychiatric referral; (2) serves as a form of treatment by bringing a sense of control to the patient and her family and allowing for implementation of behavioural modification, stress reduction and other salutary lifestyle changes; (3) helps the clinician to tailor treatment for symptom type, severity and duration; (4) assists with selection of optimal therapeutic approach, e.g. need for daily versus luteal therapy with serotonergic agents or suitability of hormonal treatments. In summary, clinical diagnosis is one of exclusion and relies on of subjective daily recording of symptoms, confirmation of premenstrual cyclicity, and exclusion or exacerbation of other disorders with similar symptoms.

The current treatments of PMS/PMDD reflect the state of knowledge regarding the pathophysiology of the disorder and the available contemporary pharmacotherapy (27–30). Susceptible women become symptomatic with the neuro-endocrine fluctuations triggered by ovulation. Logical ways of treating PMS/PMDD would be elimination of the rise and fall of the ovarian activity that triggers the PMS symptoms. Serotonin deficiency has been implicated and there are multiple pharmacological agents that augment serotonergic functioning. Modulation of the GABA_A subunit configurational changes afforded by progesterone metabolite exposure has yet to be attempted. Alternatively methods to reduce the neuro-endocrinological susceptibility through dietary and environmental interventions can also be effective. It is prudent to begin therapy with approaches with the fewest side effects, but with some efficacy based on randomized controlled trials (RCTs).

During the 2 months when the patient will be completing the prospective daily ratings it is reasonable to initiate behavioural and/ or dietary nonpharmacological interventions (27) that are healthful and at least, in part, evidence-based including any of the following:

Initiation of or increase in aerobic exercise can reduce premenstrual mood and physical symptoms of PMS. Controlled trials are lacking, but exercise has other obvious health advantages. At least 3 days per week with 20–30 min of aerobic exercise should be recommended.

Calcium deficient diets are associated with increased PMS symptoms in adolescents. Supplementation with calcium has been demonstrated to be superior to placebo for PMS symptoms in two double-blind RCTs. Calcium carbonate 1200 mg/day in divided doses can be initiated. A meta-analysis of RCTS of daily vitamin B6, a cofactor in the synthesis of 5-HT concluded that doses up to 100 mg/day are likely to be modestly helpful in the treatment of premenstrual symptoms, including depression. Caution patients to avoid higher doses as even 200 mg per day can cause neuropathy.

CBT can also be initiated during the daily recording months or as an initial treatment, especially for those who identify stress as an exacerbating factor or for adolescents with less severe PMS or for whom pharmacotherapy may be less desirable. There is evidence for reduction in PMS symptom severity with CBT. Additionally, CBT added to the SSRI, fluoxetine demonstrated longer maintenance of treatment effect than fluoxetine alone, but combining CBT and fluoxetine did not confer added benefit in terms of degree or rate of response.

NSAIDS are useful for physical pain symptoms but do not improve mood.

Upon review of the daily ratings, the clinician may find significant mood symptoms do not end with the cessation of menses, but persist into the mid-follicular phase, indicating an underlying mood or anxiety disorder. This should prompt referral to a psychiatrist or psychologist. Clinicians familiar with the treatment of mood disorders may elect to initiate daily antidepressants, and to monitor the patient closely for response and side effects. Confirmation of a PMS or PMDD diagnosis that hasn’t responded to the above measures requires pharmacologic intervention (Box 8.1.4.4).

Hormonal contraceptives containing both an oestrogen and a progestin (combined oral contraceptive) is the most commonly used reversible birth control method worldwide. Non-contraceptive benefits of oral contraceptives are significant and include cycle regulation, reduction in anaemia, functional ovarian cysts, acne, dysmenorrhoea, and pelvic pain related to endometriosis, and risk reduction for ovarian and endometrial carcinoma. There is now also evidence that certain oral contraceptives can be beneficial for PMS and PMDD; however, most oral contraceptives and progestin-only methods have not shown efficacy for this indication and currently cannot be recommended. Historically, oral contraceptives were not found in RCTs to be beneficial for women with PMS, despite elimination of ovulation. The dose of ethinyl oestradiol, the chemistry of the 19 nor-testosterone derived progestins, and the 7-day hormone-free interval in the various pill formulations have been hypothesized to provoke PMS-like symptoms. The 7 days of placebo pills may also allow for follicular development and a mini-cycle of exposure to and withdrawal from endogenous steroids, and residual PMS-like symptoms in susceptible women (28).

A newer oral contraceptive, containing 20 µg of ethinyl oestradiol and 3 mg of a novel progestin, drospirenone in a 24/4 regimen has demonstrated efficacy in approximately 60% of subjects for the treatment of PMDD, and was significantly better than placebo in two pivotal RCTs, resulting in a FDA indication for the treatment of PMDD for women who desire hormonal contraception (29–31) Studies in women with less severe symptoms of PMS, trials lasting for more than three cycles, and comparative ‘head to head’ oral contraceptive trials are lacking (30). Drospirenone, a 17–α spirolactone derivative, is chemically similar to the antihypertensive diuretic, spironolactone. Like the parent compound, drospirenone has anti-androgenic and anti-mineralocorticoid activity. The 20 µg ethinyl oestradiol/3 mg drospirenone oral contraceptive formulation in a 24-day active/4-day placebo pill cycle ensures more complete hormonal suppression. Shortening the hormone-free interval from 7 to 4 or fewer days serves to maintain sufficient circulating levels of exogenous oestrogen and progestin to better inhibit follicular development and suppress ovarian steroid synthesis. The long half-life of drospirenone (30 h) insures that there is still a presence into the 4-day hormone-free interval. The anti-mineralocorticoid and anti-androgenic action combined with the ovulation inhibition, and the lowered progestin fluctuation may contribute to efficacy. Nausea, headache, and breakthrough bleeding episodes were more prevalent in the oral contraceptive users than in those assigned to placebo, but were not higher than seen with other low dose oral contraceptives. Other extended regimen/continuous low dose oral contraceptive regimens may also be useful for PMS, but data are lacking.

For women with a formal diagnosis of severe PMS or PMDD, these agents are also first line therapy (31–33). For patients who do not desire hormonal contraception, have residual symptoms on oral contraceptives, or have contraindications for oral contraceptives, SSRIs are the treatment of choice. It is important to rule out bipolar disorder with a medical history and the aide of the daily rating scales, as SSRIs can trigger a manic episode if prescribed without a mood-stabilizing agent. In the case of severe premenstrual depression with episodes of activation, agitation, or prolonged sleeplessness, suicidal ideation, or the adolescent patient; the management should include a referral to a psychiatrist.

Over 20 RCTs and two meta-analyses have demonstrated a 50–70% response rate at standard daily SSRI doses, with significant improvement compared to placebo with all currently available SSRIs and one SNRI (20, 30–33) Specifically, recommended doses are fluoxetine 20–40 mg/day, sertraline 50–150 mg/day, paroxetine 10–20 mg or 12.5–25 mg of paroxetine continuous release (paroxetine CR) per day, and citalopram 10–20 mg/day. Smaller studies with the SNRI, venlafaxine also demonstrated efficacy within the dose range of 50–130 mg daily (31). Relief of physical symptoms may require higher SSRI doses within the dose ranges listed above. No one SSRI is demonstrably better than another. Fluoxetine, sertraline, and paroxetine are the most well studied for PMDD and all three have a US FDA indication for PMDD.

Luteal phase SSRI treatment has been studied for all of the SSRIs and is also effective. Intermittent, luteal phase dosing regimens may be less effective than continuous dosing regimens. Unlike treatment of affective disorders, where a clinical response requires 4–6 weeks of exposure to an SSRI, the response in PMS/PMDD occurs within the first days of treatment. Withdrawal symptoms are not a problem even for the shortest-acting agent, paroxetine. Luteal phase dosing should begin about 14 days before expected menses and continue until the onset of menstrual flow. Symptom onset dosing, i.e. tailoring the duration of therapy to the duration of symptoms and starting the medication at that time, can also be effective, but has not been well studied. Severe PMDD may be less likely to improve with luteal or symptom onset dosing.

Since both continuous and intermittent dosing regimens are effective, and the decision as to which regimen to use has to be made on an individual basis. Continuous dosing is indicated for women with a history of comorbid anxiety or depressive disorders or for women who experience concurrent subsyndromal anxiety or mood symptoms throughout their cycle. These women may also benefit from semi-intermittent dosing, i.e. continuous therapy with a dose increase during the late luteal phase. Treatment with an agent that is effective for both depressive disorders and PMDD is warranted for women with both diagnoses. Similarly, premenstrual worsening of mood symptoms in a woman with major depressive disorder may require increased premenstrual dosing of a serotonergic agent or addition of a serotonergic medication to an otherwise effective ongoing psychotropic regimen. Continuous dosing is preferable for women who find it easier to adhere to a daily regimen. Intermittent, or even symptom-onset dosing, is appropriate for women with ‘pure’ PMS/PMDD, i.e. those whose symptoms are limited to the late luteal phase only, with a definite ‘on-off’ presentation. This should also be the option for women who prefer not to take medication throughout the entire cycle or who experience bothersome side effects (especially sexual dysfunction) which can be minimized with intermittent dosing.

After 1–2 cycles, treatment response and side effects should be assessed. If response is suboptimal or there are intolerable side effects, dose may be altered or another SSRI should be tried, since the various SSRIs have somewhat different mechanism of action. SSRIs should be taken in the morning unless sedation occurs, in which case the dose can be taken in the late afternoon, generally before 16.00 h to prevent insomnia. Gastrointestinal side effects, anxiety, and headache are usually transients; weight gain can also occur. The most common side effects of SSRIs when used to treat PMDD included anxiety, insomnia, and nausea. All SSRIs and even SNRIs can be associated with decreased libido, delayed orgasm, and anorgasmia in a small percentage of cases. Besides lowering doses or duration of serotonergic therapy, antidotes for sexual side effects include adding a dopaminergic agent, such as buproprion, adrenergic agents, such as methylphenidate; adding the 5-HT_1A agonist, buspirone; or trying a phosphodiesterase inhibitor, like sildenafil. SSRIs can be used safely with oral contraceptives without loss of efficacy of either class of agents.

Although these medications reduce symptoms, patient compliance must be considered when evaluating treatment. A Cochrane review of SSRIs as treatment for premenstrual disorders revealed that patients taking SSRIs are 2.5 times more likely than women using placebo to withdraw from a study due to side effects especially sexual dysfunction and weight gain (20). Additionally, many women never fill the SSRI prescription due to the stigma of taking a psychiatric medication or fear of addiction. Another study found a discontinuation rate of over 60% after 2 years.

Other psychotropic medications with some efficacy include luteal buspirone 10 mg two to three times per day or alpazolam, 0.25–0.5 mg taken up to two to three times per day, limited to the symptomatic days (31, 33). Alprozolam or other benzodiazepines should only be administered in the luteal phase due to the potential for addiction and tolerance. Although these agents have shown only modest usefulness for PMS/PMDD compared to SSRIs, their intake results in fewer sexual side effects. These agents or other anxiolytics can be added to an SSRI or oral contraceptive in the luteal phase if anxiety persists with the latter agents alone.

Depending on the severity of specific physical symptoms, these symptoms may need to be addressed separately. In women with severe mastalgia and bloating, use of spironolactone, an aldosterone receptor antagonist, may be a helpful adjunct and has a few small RCTs to support use (31). A dose of 100 mg from day 12 of the cycle to the onset of menses can result in significant improvement in abdominal bloating, swelling of extremities, breast discomfort, and even mood symptoms of irritability, depression, and anxiety. It should be noted that the drospirenone in the EE 20 µg/drospirenone 3 mg 24/4 regimen oral contraceptive is equivalent to only 25 mg of spironolactone.

GnRH agonist to suppress gonadotropins and ovarian sex steroids provides symptomatic relief in the majority of women with PMS, even with hormone add-back (25, 29); however, side effects of the hypo-oestrogenic state generally preclude use without hormone ‘add-back’. A menopausal dose of continuous oestrogen/progestin or tibilone as hormone ‘add back’ can prevent vasomotor symptoms and vaginal atrophy, and help maintain bone density and cardiovascular health, but even with the addition of oestrogen and progestin, there is concern about long-term consequences. GnRH analogue remains a third-line therapy, indicated if all other approaches have failed or if there is another concurrent indication for use, such as cyclic pelvic pain or endometriosis. GnRH analogues are a useful diagnostic tool prior to contemplating definitive surgical treatment with bilateral salpino-opherectomy (BSO) or when determining if severe premenstrual symptoms are related to PMDD or a co-morbid affective disorder. Other side effects of GnRH analogues include myalgias, arthralgias, headaches, and urogenital atrophy.

Danazol, an androgen analogue is another option for women who have PMS with concurrent endometriosis or severe mastalgia. Side effects are substantial and are dose-dependent: weight gain, acne, hirsutism, and decreased breast size. Doses of 100–200 mg twice per day can be tried (29, 31). It does not provide reliable contraception and many women will ovulate with the 400 mg daily dose. Danazol can virilize a developing fetus, therefore, double barrier, IUD, or sterilization as contraception (with discussion of abortion if a pregnancy does ensue) is imperative in sexually active patients. Those who ovulate are not likely to have a positive treatment response.

Transdermal oestradiol (200 µg) has been demonstrated to effectively control PMS symptoms as ovulation is inhibited with the high oestradiol dose. If this modality is used, the endometrium must be protected with a progestin. Since progestins can recreate PMS symptoms, an intrauterine device containing a progestin is preferable to protect the endometrium while keeping blood levels (and therefore CNS levels) of progestogen low. Although there is some evidence that high dose oestradiol is effective in treating PMS and the levonorgestrel containing IUD prevents, or even reverses endometrial hyperplasia, there are no published data to confirm the efficacy of the combination (29, 31).

Tibolone is a synthetic steroid with oestrogenic, androgenic, and progestogenic properties, with many studies demonstrating efficacy for menopausal symptoms. There is some evidence from one small randomized placebo-controlled cross-over trial that 2.5 mg daily may me effective treatment for PMS. Further confirmatory trials are necessary before recommendations for this treatment can be made, however, it may be useful as add-back during GnRH therapy.

Castration is effective for severe PMS/PMDD and may be reasonable option for patients in their 40s who have completed childbearing and have failed other medical treatments, or who are planning a hysterectomy for other gynecologic indications. Hysterectomy alone is not sufficient to effectively relieve PMS symptoms, and BSO alone requires both progestin and oestrogen replacement, which leaves the risk of re-stimulating symptoms. Castration before age 60 may increase various causes of mortality especially cardiovascular.

CAM treatments have shown some efficacy in treating premenstrual disorders, however, results are mixed and most studies are either small or not well controlled. Some of these alternative therapies include: bright light therapy, herbal, and nutritional supplements, such as Chaste tree extract, and various mind-body approaches.

Twenty percent of reproductive aged women suffer from PMS and up to 8% from PMDD, a severe form of PMS with predominant affective symptoms. Once established, the syndromes persist for over most of the reproductive years, and are responsible for substantial impairment in relationships and productivity. The aetiology of the disorders is still enigmatic and likely multifactorial; disorders of the serotonergic and GABAergic systems have been proposed. Once a definitive diagnosis is made with the aide of history, physical examination, and prospective symptom charting, initiation of various treatments can be effective. Education, lifestyle changes (exercise, and stress reduction), cognitive behavioural therapy, and calcium supplementation can be utilized during the diagnostic phase. The oral contraceptive containing 20 µg of ethinyl oestradiol and 3 mg drospirenone in a 24/4 preparation has shown efficacy in RCTs and received an FDA indication for PMDD for women who also desire hormonal contraception. UP to this time, other oral contraceptives have not demonstrated efficacy. The administration of a serotonergic antidepressant in the luteal phase or every day is also a first line strategy and likely to alleviate physiological and physical symptoms, and improve quality of life and functional status. Luteal phase dosing is the preferred method of treatment with SSRIs in most cases, as cost and overall side effects are minimized. Three SSRIs have FDA indications for PMDD. The diuretic, spironolactone can be effective, particularly for physical symptoms. GnRH analogues with ‘hormonal add-back’, danazol, high-dose oestradiol patches with progestin, or bilateral oophrectomy are also in the armamentarium and may be indicated in very select cases.