Premenstrual Disorders: Epidemiology and Disease Burden

December 1, 2005
Daniel R. Mishell, Jr., MD
Daniel R. Mishell, Jr., MD

Volume 11, Issue 16 Suppl

Many women of childbearing age experience avariety of symptoms related to the menstrual cyclethat may be limited to mild discomfort or extend topremenstrual syndrome or, depending on the degreeof emotional and somatic impairment, to the mostsevere premenstrual dysphoric disorder (PMDD).With PMDD, women experience markedly compromisedquality of life and ability to function in severalsettings, leading to higher direct medical costs forincreased physician visits and laboratory tests, andhigher indirect costs to employers through lower productivityat work. Diagnostic criteria established forpremenstrual disorders may now enable physiciansto evaluate their patients' complex symptoms moreprecisely and recommend a form of treatment thatcan alleviate the original complaint and improvehealthcare for women.

(Am J Manag Care. 2005;11:S473-


During a woman's reproductive years,premenstrual symptoms can significantlydisrupt her quality of life(QOL) as well as present difficult diagnosticand management challenges for clinicians.Premenstrual disorders primarily occur inwomen who have ovulatory menstrual cycles;however, premenstrual syndrome (PMS)symptoms have also been described in oligoovulatorywomen, for example, in womenwith polycystic ovary syndrome and those inperimenopause, who may have bleeding withoutovulation. Also, some women on oral contraceptiveshave persistent PMS symptoms.The disturbing symptoms of PMS or premenstrualdysphoric disorder (PMDD) occur duringthe luteal (premenstrual) phase ofovulatory cycles (the last 14 days before menstruation)and disappear shortly after theonset of menstruation. Normal, manageablesymptoms associated with premenstrualmolimina (signs of impending menstruation)that most ovulatory women experienceinclude breast tenderness, food cravings, andbloating or pelvic heaviness.1 These symptomsdo not adversely affect a woman's abilityto function during daily activities. PMS ischaracterized by disturbing physical and/ormood-related symptoms that occur duringthe late luteal phase of the menstrual cycleand usually disappear within 3 days after theonset of menses. The most severe type of premenstrualdisorder, PMDD, formerly knownas late luteal-phase dysphoric disorder, causessignificant functional impairment andgreatly diminishes a woman's QOL.

Development of Diagnostic Criteria forPMS and PMDD

Premenstrual symptoms are characterizedin terms of their type, severity, and timing.Until relatively recently, PMS researchwas limited by the lack of a clinically meaningfuldefinition for this disorder. Clinicalmanagement guidelines for PMS that includediagnostic criteria were published in 2000by the American College of Obstetriciansand Gynecologists (ACOG).2 These guidelineswere based on an earlier publication byMortola and colleagues, who first noted thedistinction between endogenous depressionand depressive episodes occurring onlyduring the luteal phase of cycles of womenwith PMS.3

ACOG practice guidelines for a diagnosisof PMS specified that one or more disturbingaffective or somatic symptoms (Table 1)must have occurred during the 5 days beforemenses in each of 3 previous menstrualcycles. These symptoms must be relievedwithin 4 days of the onset of menses andmust not recur until at least day 13 of thewoman's cycle.2 The symptom most oftenassociated with PMS is irritability.

For a diagnosis of PMS, a woman whoexperiences these symptoms must sufferfrom notably impaired function in her socialactivities or work-related performance. Inaddition, her symptoms must occur consistentlyduring 2 cycles of prospective recordingand in the absence of pharmacotherapy,hormone administration, or use of drugs oralcohol.

Diagnostic and Statistical

Manual of Mental Disorders, Fourth

Edition, Text Revision (DSM-IV-TR)

For PMDD, the American PsychiatricAssociation has established criteria for diagnosisin the .4 Awoman must experience at least 5 of thesymptoms listed in Table 2, including 1 ormore of the core symptoms, to have PMDD.These symptoms must have been presentduring most of the last week of the lutealphase in a majority, if not all, of the woman'smenstrual cycles during the previous year.Also, symptoms must begin to remit within afew days of the start of the follicular phase ofher cycle, and they must not recur duringthe week after menses.

Collectively, the symptoms must be severeenough to disrupt a woman's ability to functionat work or school, during household orsocial activities, or in interpersonal relationships.However, they must not beindicative merely of an exacerbation of ongoinganxiety or depressive disorders, withwhich they may share common features.Prospective daily recording of symptomsduring at least 2 consecutive symptomaticcycles is necessary for confirmation of thediagnostic criteria. Because of the cyclicnature of depressive symptoms in PMDD, itshould be possible to differentiate this premenstrualdisorder from depressive disordersthat occur throughout the cycle.4

Epidemiology of Premenstrual Disorders

Approximately 70% to 90% of women ofchildbearing age in the United States experienceat least some uncomfortable symptomsduring the premenstrual phase of theircycles; this percentage can be extrapolatedto approximately 43 million to 55 millionwomen (annual estimates).5 Between 20%and 40% of this group, or approximately 12million to 25 million women (annual estimates),believe that they have symptomssufficiently bothersome to qualify as PMS.5In addition, an estimated 3% to 8% of USreproductive-age women, numbering 2million to 5 million (annual estimates), havesymptoms of sufficient severity to be classifiedas PMDD.5

Premenstrual disorders can affect awoman at any stage in her reproductivelife—beginning around age 14, or about 2years after menarche, and persist untilaround age 51, when menopause typicallyoccurs.6 According to Halbreich and colleagues,US women have approximately 481menstrual cycles during this age span.6 Withan average adjustment of 22 months for 2pregnancies and postpartum periods, manywomen actually experience about 459 cyclesduring their childbearing years. In addition,US women with PMDD experience an averageof 6.4 days of severe symptoms per menstrualcycle7; this extrapolates to a lifetimetotal of roughly 2938 days, equivalent toapproximately 8 years with severe symptoms.Thus PMS/PMDD can cause distress oreven impairment of functioning over a significantfraction of a woman's lifetime. Evenafter treatment that alleviates the symptomsof PMDD, the underlying physiologic abnormalityremains, so that symptoms may recurafter discontinuation of treatment.8

In addition to the projections listed earlierfor numbers of women with PMS/PMDD,recently published studies regarding diagnosticand treatment approaches for womenwith premenstrual disorders were evaluated.6 For example, Kraemer and Kraemeranalyzed the experience of women whoreported seeking medical attention for theirpremenstrual symptoms from an average of3.75 physicians over an average of 5.33years before being diagnosed with PMS.9Similarly, 85% of women respondents to asurvey by Campbell and colleagues had tried1 or more treatments for PMS, and 45%reported that they wanted additional help.10In a survey of 445 US women reporting premenstrualsymptoms, 138 (31%) met the criteriafor PMS.7 Less than half (45%) of thegroup with severe premenstrual symptomshad sought medical treatment, and 58% didnot think any treatment would help.7 Itappears likely that many women with PMSand PMDD may be misdiagnosed, becausesymptoms can mimic organic disorders orworsen existing psychiatric conditions(known as premenstrual magnification), andoften they may not receive adequate counselingto assuage their concerns.

In an investigation of pretreatment psychosocialfunctioning of women with PMDDand their response to treatment, scores onseveral psychosocial functioning scales duringthe follicular phase in these women weresimilar to community norms, whereas scoresduring the luteal phase were similar to scoresof women with depressive disorders.11

The spectrum of premenstrual symptomseverity was examined by Sternfeld et al in alarge, diverse sample of women from membersof a large health maintenance organizationin northern California.12 Eligibilitycriteria included: (1) aged 21 to 45, (2) nopsychiatric diagnoses in the past 2 years, (3)no psychotropic medication in the past 6months, (4) no history of hysterectomy orbilateral oophorectomy, and (5) no pregnancyin the previous year. The 1194 womenwho qualified for the study rated their symptoms' severity during each day of the premenstrualphase of their cycles: 186 hadminimal symptoms, 801 had moderatesymptoms, 151 had severe symptoms, and56 met the criteria for PMDD. The percentageof women in each category is shown inthe Figure. Each woman experienced a relativelyconstant degree of symptom severity—particularly for emotional symptoms—over 2consecutive cycles.

Premenstrual Disorders:The Disease Burden

The high prevalence of premenstrual disordersand their negative impact onwomen's QOL has been an ongoing concernfor patients and their physicians. Becausethe symptoms are so varied, and because nospecific endocrine diagnostic test exists, apremenstrual disorder may go unrecognizedor be misdiagnosed as another condition.However, diagnostic criteria developed forPMS and PMDD have made it possible toconsider more objectively the somatic,emotional, and behavioral symptoms thatcause long-term distress for so manywomen. Overcoming this barrier to understandingthe complexity of women's menstrual-related complaints is a major steptoward ameliorating patients' impaireddaily functioning and achieving adherencewith physician recommendations.

In addition to interfering with a woman'sQOL, PMS and PMDD can have both directand indirect economic consequences. Directcosts take the form of fees for outpatientoffice visits (including possible referrals tospecialists), laboratory tests, and treatment.Indirect costs, which are considerably moredifficult to quantify, are usually viewed interms of lost productivity at work (sometimestermed "presenteeism") and lostwages because of absenteeism.




Borenstein and colleagues analyzedwomen aged 18 to 45 years who were enrolledin a medical group in southernCalifornia in a cross-sectional cohortstudy.13 Women who completed the screeningsurvey and who met eligibility criteriawere asked to complete 2 screening toolsin a telephone interview: the MedicalOutcomes Study Short Form-36 (SF-36)and the 10-item Center for EpidemiologicalStudies-Depression Scale. Of the 436 womenwho completed and returned the survey, 125were identified as having PMS and 311served as controls. The women with PMSscored significantly lower on the mental (<.001) and physical (= .04) scales of theSF-36 in comparison with controls. In addition,women with PMS reported decreasedproductivity at work, greater interferencewith hobbies, and a greater number of workdaysmissed for health-related reasons (each<.001) compared with the control group.Women with PMS also made more frequentvisits to ambulatory-care providers andwere more likely to accrue an excess of$500 in visit costs over 2 years. It was concludedthat PMS has a significant impact ona woman's health-related QOL and maylead to reduced productivity at work andincreased healthcare costs.13


This same group of women was also askedto maintain a symptom diary using the DailyRecord of Severity of Problems (DRSP) during2 consecutive menstrual cycles.14 A totalof 125 participants were identified as havingPMS; of these, 78 women had symptoms for 1cycle and 47 women had symptoms for 2cycles; the 311 women without PMS againserved as controls. Table 3 shows the percentagesof controls and women with 2 cycles ofPMS who had high health-related absenteeism(>2 workdays/month missed), highproductivity loss (>5 days/month with a50% decrease in productivity), and a highdegree of impairment (>14 days/month) inoccupational and social activities andother settings. The women with 2 cyclesof PMS were significantly more likely toexperience general impairment on moredays per month compared with controls(22 &#177; 6.5 days/month vs 9.6 &#177; 7.9 days/month; <.0001).



In conjunction with the preceding study,Borenstein and colleagues also quantifiedthe economic impact of PMS on the employerby looking at direct medical costs (basedon administrative claims and the Medicarefee schedule) and indirect costs (based onthe woman's self-report of workdays missedand decreased work productivity).14 Thepopulation was composed of 374 womenwith regular menses who had used the DRSPto record daily symptoms for 2 consecutivemenstrual cycles. Based on the DRSP, 111(29.6%) women were diagnosed with PMS.This group had increased direct medicalcosts of $59 per year (= .003) andincreased indirect costs of $4333 per yearcompared with women who did not havePMS (<.0001). It was concluded that adiagnosis of PMS was associated with modestincreases in direct medical costs and a considerableincrease in indirect costs stemmingfrom missed workdays and lowerproductivity when the women were atwork.14

Etiology of Symptoms inPremenstrual Disorders

Work by a number of investigators suggeststhat reproductive hormones are normalin women with PMS/PMDD, but womenwith these syndromes have higher sensitivityto the changing levels of reproductive hormonesthat occur during the menstrualcycle. Recently, Halbreich and Monacelli15reviewed concepts on the pathobiology ofpremenstrual disorders and noted that thesedisorders involve multifaceted interactionsbetween processes of the central nervoussystem, hormones, and other modulators.Women who have a genetic predisposition topremenstrual disorders (although no specificgenes have yet been identified) may experienceabnormally severe responses tonormal cyclic fluctuations of gonadal hormonesthat may contribute to symptomexpression.

Ovulation and gonadal hormones canmediate changes in neurotransmitters, suchas serotonin and gamma-aminobutyric acid,which are involved in regulating mood,behavior, and cognitive functions,16 and inneurohormonal systems, such as the reninangiotensin-aldosterone system (RAAS),which controls sodium and water retentionand potassium excretion, and any resultingabnormalities can lead to premenstrualsymptoms. For example, estrogen and progesteroneaffect the RAAS in different waysto influence electrolyte and fluid balance inthe body. Estrogen has a mineralocorticoideffect by inducing the synthesis of angiotensinogenin the liver, and this, in turn,increases aldosterone and stimulates theRAAS to increase fluid retention, bloating,and breast tenderness—all premenstruallyrelated symptoms. Progesterone, however,exhibits antimineralocorticoid activity bycompeting with aldosterone at the aldosteronereceptor. The end result of these hormonalchanges is an increase in fluidexcretion and a reduction in the bloatingand breast tenderness that frequently occurin the late luteal phase of the menstrualcycle.13 Therefore, a woman who experiencessignificant water retention, bloating,and breast tenderness in the late lutealphase of her cycle has an altered responseto changes in levels of gonadal hormonesthat contributes to her premenstrualsymptoms.16

Another study evaluated the episodicrelease of progesterone and luteinizing hormone(LH) in the luteal phase of the cyclesof 14 women with PMS and in 14 controlswithout PMS.17 The Moos' Menstrual DistressQuestionnaire was used to prospectivelyconfirm PMS in 2 consecutive menstrualcycles. Blood samples were drawn every 10minutes for 12 hours, and the Detect programwas used to determine the presence ofsignificant progesterone and LH pulses. Theincreased pulse frequency and decreasedamplitude of LH in women with PMS comparedwith controls reflects changes in neurotransmitters,supporting the concept ofPMS as a neuroendocrine disorder.17

Diagnosis of PMS and PMDD

To establish a diagnosis of PMS or PMDD,symptoms must: (1) be characteristic ofcommon mood or physical symptoms, cognitivedisturbances, and/or behavioral consequences;(2) be limited to the luteal phaseof the cycle; (3) cause discomfort for thewoman and, in the case of PMDD, greaterimpairment; and (4) not be readily explainedby another diagnosis.17

During the initial visit of a woman presentingwith premenstrual symptoms, severalissues should be addressed. Based onthe patient's history and a physical examination,the physician should determinewhether 1 or more other diagnoses shouldbe considered and, if indicated, shouldorder appropriate laboratory tests or consultationsto assess any other possibilities.If the patient is experiencing vegetativesymptoms, significant suicidal ideations,or frequent inability to function, a psychiatricor psychologic referral should beconsidered.18

At least 2 months of prospective dailysymptom recording using an establishedassessment tool provided by the physicianare required for a diagnosis of PMS or PMDD.Examples of tools to evaluate premenstrualsymptoms appear in Table 4.

At the second visit, the physician shouldcritically review the pattern of the patient'ssymptoms in her daily record. For example,is the menstrual pattern consistentwith ovulation? Are the symptoms confinedto the luteal and early menstrualphases of the cycle? If the assessment tooland the overall clinical evaluation are consistentwith a premenstrual disorder, theclinician should then select an appropriateform of treatment.18

Additional daily symptom recording canhelp the physician and patient assessresponse to recommended treatments. Forexample, if symptoms are relatively mild,lifestyle modifications, dietary and othernutritional approaches, and nonprescriptionremedies may be tried and their effects documentedover several cycles. Not all patientsare willing and able to invest the timerequired for long-term symptom recording,however.


Most women of childbearing age experiencesome cyclic menstrually relatedsymptoms, the severity of which can rangefrom being relatively mildly problematicto causing serious premenstrual distress andinterfering with occupational and socialfunctioning. Across their entire constellation,premenstrual symptoms have beenshown to have an adverse impact on awoman's QOL and productivity and to leadto increased direct and indirect medicalcosts.

The proposed etiology of premenstrualdisorders is multifaceted and includes anunderlying genetic predisposition (althoughno genes have yet been identified)that makes a woman more susceptible tochanges in gonadal hormones that interactwith neurotransmitters and neurohormonalsystems, resulting in symptoms thatoccur only in the luteal phase of the menstrualcycle.16

Diagnostic criteria have been developedfor the diagnosis of PMDD that require (1) atleast a 30% increase in at least 5 symptomsfrom the follicular to the luteal phase over 2menstrual cycles based on daily prospectiverecords; (2) the occurrence of menstruallyrelated symptoms that cause dysfunction orimpairment; (3) confirmation that symptomsoccur only premenstrually; and (4) laboratorytesting and/or consultations torule out or identify other likely causes forthe symptoms. These objective measuresincrease the likelihood of evaluating patientsymptoms more accurately on presentationand recommending treatment approaches toalleviate symptoms, restore function, andoptimize overall health for women with premenstrualdisorders.

Address correspondence to: Daniel R. Mishell, Jr, MD, Women andChildren's Hospital, 240 N. Mission, Room L1009, Los Angeles, CA 90033;

Obstet Gynecol Clin North Am.

1. Kessel B. Premenstrual syndrome. Advances in diagnosisand treatment. 2000;27:625-639.

Obstet Gynecol.

2. ACOG Practice Bulletin. Premenstrual syndrome.2000;95:end of issue 4.

Am J Obstet Gynecol.

3. Mortola JF, Girton L, Yen SS. Depressive episodes inpremenstrual syndrome. 1989;161:1682-1687.

Diagnostic and

Statistical Manual of Mental Disorders, Fourth Edition,

Text Revision.

4. Premenstrual dysphoric disorder. In: Washington, DC: American PsychiatricAssociation; 2000:771-774.

Practical Strategies in Obstetrics and


5. Ginsburg KA, Dinsay R. Premenstrual syndrome. In:Ransom SB, ed. Philadelphia, Pa: WB Saunders; 2000:684-694.


6. Halbreich U, Borenstein J, Pearlstein T, Kahn LS. Theprevalence, impairment, impact, and burden of premenstrualdysphoric disorder (PMS/PMDD). 2003;28(suppl 3):1-23.

J Womens Health Gend

Based Med.

7. Hylan TR, Sundell K, Judge R. The impact of premenstrualsymptomatology on functioning and treatmentseekingbehavior: experience from the United States,United Kingdom, and France. 1999;8:1043-1052.

Am J Obstet Gynecol.

8. Pearlstein T, Joliat MJ, Brown EB, Miner CM.Recurrence of symptoms of premenstrual dysphoric disorderafter the cessation of luteal-phase fluoxetine treatment.2003;188:887-895.

J Womens Health.

9. Kraemer GR, Kraemer RR. Premenstrual syndrome:diagnosis and treatment experiences. 1998;7:893-907.

J Reprod Med.

10. Campbell EM, Peterkin D, O'Grady K, Sanson-Fisher R. Premenstrual symptoms in general practicepatients. Prevalence and treatment. 1997;42:637-646.

J Clin Psychiatry.

11. Pearlstein TB, Halbreich U, Batzar ED, Brown CS,et al. Psychosocial functioning in women with premenstrualdysphoric disorder before and after treatment withsertraline or placebo. 2000;61:101-109.

Obstet Gynecol.

12. Sternfeld B, Swindle R, Chawla A, Long S, KennedyS. Severity of premenstrual symptoms in a health maintenanceorganization population. 2002;99:1014-1024.

J Reprod Med.

13. Borenstein JE, Dean BB, Endicott J, et al. Healthand economic impact of the premenstrual syndrome.2003;48:515-524.

J Occup Environ Med.

14. Borenstein J, Chiou CF, Dean B, Wong J, Wade S.Estimating direct and indirect costs of premenstrual syndrome.2005;47:26-33.

Prim Psychiatry.

15. Halbreich U, Monacelli E. Some clues to the etiologyof premenstrual syndrome/premenstrual dysphoricdisorder. 2004;11:33-40.


16. Halbreich U. The etiology, biology, and evolvingpathology of premenstrual syndromes. 2003;28(suppl 3):55-99.

J Clin Endocrinol Metab.

17. Facchinetti F, Genazzani AD, Martignoni E, FioroniL, Nappi G, Genazzani AR. Neuroendocrine changes inluteal function in patients with premenstrual syndrome.1993;76:1123-1127.

Obstet Gynecol.

18. Johnson SR. Premenstrual syndrome, premenstrualdysphoric disorder, and beyond: a clinical primer forpractitioners. 2004;104:845-859.