Introduction

Premenstrual dysphoric disorder (PMDD) is commonly described as a more severe form of premenstrual syndrome (PMS). According to the Mood Disorder Work Group (hereafter, the Work Group) for the American Psychiatric Association’s fifth edition of its Diagnostic and Statistical Manual of Mental Disorders (DSM-5), PMDD affects 2 percent to 5 percent of premenopausal women (Epperson et al. 2012), in contrast to PMS, which is said to affect anywhere between 30 to 40 percent (Pearlstein 2007) and 75 percent (Steiner 2000) of menstruating women. Symptoms of PMDD include marked depressed mood, anxiety, “affective liability,” irritability/anger, change in appetite, reduced interest in usual activities, hypersomnia or insomnia, difficulty concentrating, feeling “out of control,” loss of energy, and other physical symptoms such as breast tenderness, bloating, and weight gain (American Psychiatric Association 2013a). These symptoms, while similar to those of major depressive disorder, should be cyclical and begin to improve once menstruation begins (American Psychiatric Association 2013b).

Previously, PMDD was listed with its own code in the appendix (diagnoses requiring further research) of the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) and in the main text under “depressive disorder not otherwise specified.” With the release of the DSM-5 in May 2013, PMDD has now been upgraded to a full category in the main text. (Judging by the beta draft of the upcoming edition of the World Health Organization’s International Statistical Classification of Diseases and Related Health Problems, PMDD also appears set for inclusion as a separate disorder in the ICD-11, due to be published in 2017 [World Health Organization 2013].) As the main text is purportedly reserved for scientifically validated categories, the move sends a strong signal to researchers and clinicians of the legitimacy of the diagnosis. In a world where the medical narrative trumps all other narratives (Frank 1995), the significance of this move cannot be overstated. The DSM’s influence extends far and wide, particularly in North America (Lafrance and McKenzie-Mohr 2013). From guiding clinical practice and providing a framework for many psychiatric textbooks as well as textbooks in other related fields (Cosgrove et al. 2006; Marecek and Hare-Mustin 2009) to influencing the pharmaceutical industry, insurance, law, government, art, and the media (Kirk and Kutchins 1992; Ussher 2010), the DSM has retained its influential power despite criticism (Pilgrim 2007). In the words of Hare-Mustin and Marecek, “the DSM has become the pre-eminent organizing rubric” (1997, 112). Thus, when a diagnosis such as PMDD is upgraded in the DSM, people notice and take it seriously.

The move met with some opposition, similar to that when PMDD was first included in the DSM-IV (King 1990). The opposition has come mainly from feminists who worry about the effect that diagnostic and treatment approaches to mental health issues have on oppressed groups (e.g., Caplan and Cosgrove 2004). In classifying certain mental and behavioural states as disordered, the DSM does more than just categorise them as abnormal or problematic; it also metaphorically rubber stamps them as diseases or disorders that can, or should, be medically treated. This, in turn, can have a significant impact on those labelled with such a “disorder.” It is also worth remembering that it is the powerful who decide what is normal and what is not (Marecek and Hare-Mustin 2009). For example, individuals who do not identify with or conform to the gender norms of their sex can be diagnosed as having gender dysphoria (or gender identity disorder as it was previously known in the DSM and the ICD).Footnote 1 The APA’s fact sheet on gender dysphoria states that:

For a person to be diagnosed with gender dysphoria, there must be a marked difference between the individual’s expressed/experienced gender and the gender others would assign him or her, and it must continue for at least six months. In children, the desire to be of the other gender must be present and verbalized. This condition causes clinically significant distress or impairment in social, occupational, orother important areas of functioning.

Gender dysphoria is manifested in a variety of ways, including strong desires to be treated as the other gender or to be rid of one’s sex characteristics, or a strong conviction that one has feelings and reactions typical of the other gender (American Psychiatric Association 2013a, 1).

Although homosexuality was removed from the DSM in 1973 as a result of protests against its classification as a mental disorder, gender dysphoria arguably provides an alternative avenue to pathologise same-sex sexual desire and behaviour that conflicts with societal norms and gender roles for males and females (Newman 2002; Ault and Brzuzy 2009). “Treatment” for gender dysphoria includes the use of hormones and sex-reassignment surgery (American Psychiatric Association 2013a). While some transgendered people do seek sex-reassignment surgery, they should not have to be diagnosed with a mental illness in order to have it, nor should a genital surgery be accepted as a treatment for mental illness.

By classifying gender-nonconformism as a mental disorder, the diagnosis entrenches oppressive gender expectations and makes the fight against such prejudice even harder. As Ault and Brzuzy state: “Because of institutionalized sexism, heterosexism, and homophobia, those whose gender identities do not conform to societal expectations often experience family difficulties, discrimination, bodily harm, underemployment and social isolation” (Ault and Brzuzy 2009, 187). In this way, the DSM can further oppress groups who are already marginalised (e.g., Butler 1999; Brustow 2005; Marecek and Hare-Mustin 2009; Lafrance and McKenzie-Mohr 2013). As Duffy et al. note, “[p]eople already in tenuous positions on the social and economic ladder are the very ones most likely to be further disadvantaged by diagnosis” (2002, 372). Now that PMDD is a full category in the DSM, women, especially those already in difficult situations, may be further disadvantaged by this diagnosis. Given what is at stake, it makes sense to view the upgrade of PMDD as an ethical issue that warrants critical attention.

This paper presents the case against PMDD as a diagnosis, both summarising well-rehearsed arguments and adding new arguments as appropriate. The paper also presents some sociological arguments as alternatives to the scientific arguments in support of a biological basis for PMDD. I deliberately call them scientific arguments here rather than scientific evidence because, as Paula Caplan states, “the creation and use of psychiatric categories is rarely based on solid science, as I learned when I served on two DSM committees. The absence of science leaves a void into which every conceivable kind of bias has been found to flow—including sexism” (Caplan 2008, 63). The British Psychological Society also issued a response to the DSM-5 that highlights its concern with the insufficient evidence for the diagnostic categories in the DSM-5 as well as the negative effect that medicalising normal/natural responses to experiences has on the population—responses “which do not reflect illnesses so much as normal individual variation” (British Psychological Society 2011 2). The British Psychological Society further states:

Diagnostic systems such as these therefore fall short of the criteria for legitimate medical diagnoses. They certainly identify troubling or troubled people, but do not meet the criteria for categorisation demanded for a field of science or medicine (with a very few exceptions such as dementia.) We are also concerned that systems such as this are based on identifying problems as located within individuals. This misses the relational context of problems and the undeniable social causation of many such problems (British Psychological Society 2011, 2).

In this paper, I first present the argument that PMDD pathologises understandable anger/distress and that to do so is potentially dangerous. I then present evidence that PMDD is a culture-bound phenomenon, not a universal one. I also argue that even if (1) medication produces a desired effect, (2) there are biological correlates with premenstrual anger/distress, (3) such anger/distress seems to occur monthly, and (4) women are more likely than men to be diagnosed with affective disorders, none of these factors substantiates that premenstrual anger/distress is caused by a mental disorder. I argue that to assume they do is to ignore the now accepted role that one’s environment and psychology play in illness development, as well as arguments concerning the social construction of mental illness. In doing so, I assert that correlation does not equal causation and argue that variations from “normal” function are not necessarily pathological. Building on feminist discourse that urges us not to view women’s bodies with their particularities and cyclical nature as “abnormal” compared to men’s bodies (e.g., Ussher 1996; The Working Group for a New View of Women’s Sexual Problems 2001; Offman 2004), I present evidence that even if some women are found to be biologically predisposed to an increased sensitivity once a month, this biological difference may itself be the result of prior trauma rather than a cause (let alone the sole cause) of psychological suffering. Furthermore, an alternative view of the biological difference may even allow the possibility for it to function as an advantage rather than a disadvantage in a woman’s life. In presenting alternative explanations, I mean to show that the sociological explanations for this phenomenon are just as plausible as the biological explanations and hence that the American Psychiatric Association cannot say the question is settled in terms of the latter. Finally, I argue that there is potential danger not only from the diagnosis of PMDD, but also from its over-diagnosis.

I wish to emphasise that I am not asking for premenstrual anger/distress to be ignored, but that the diagnosis of PMDD should not be considered a mental disorder and should be removed from the DSM. To do so is the opposite of ignoring the problems that these women face. I contend that diagnosing their distress as a mental disorder ignores the underlying problems that medical treatment alone would fail to address. By treating premenstrual anger/distress as “just PMDD,” legitimate concerns can be brushed aside and we miss an opportunity to deal with what may really be at the root of their anger/distress.

Pathologising Legitimate Anger/Distress

The feminist argument is that if women are angry/distressed, it is for good reason, not due to pathology. This argument is supported by research that has found that women who are said to have PMDD are also significantly more likely than the general population to have suffered abuse, relationship problems, and mistreatment at work (Caplan 1995; Golding and Taylor 1996; Taylor 1999; Golding et al. 2000). Rates of lifetime sexual and physical abuse are greater than 60 precent in women diagnosed with PMDD (Paddison et al. 1990; Golding and Taylor 1996; Golding et al. 2000; Girdler et al. 2003). Several studies have also found premenstrual anger/distress to be related to household and childcare responsibilities (Coughlin 1990; Ussher 2003, 2004; Ussher and Perz 2010)—especially for women who do the lion’s share of the housework and childrearing in addition to having a full-time job (Ussher 2003)—and communication between couples (Kuczmierczyk, Labrum, and Johnson 1992). Further, as Stout and Steege have found, “material brought up during premenstrual outbursts often involved valid marital issues which the couple had been unable to address productively at any other time in the menstrual cycle” (1985, 627). This research, which supports the argument that if women are angry/distressed it is for good reason and not due to pathology, cannot be ignored.

Thus, the main concern is that the diagnosis of PMDD medicalises what is, at its root, a problem related to the unfair burdens and circumstances in which women can find themselves (Riessman 1983; Edwards 1984; Laws, Hey, and Eagan 1985). It is an argument that has been made not only in relation to PMDD but to almost all other human woes currently classified as mental illnesses (e.g., British Psychological Society 2011; Greenberg 2013). There is now strong evidence that the context of people’s lives plays an important part of the emotional, mental, and behavioural problems that people experience (Stoppard 2000; Bentall 2003; Read et al. 2005; Tew 2005; Wilkinson and Pickett 2009) such that the evidence for biological causes pales in comparison (Falloon 2000; Bentall 2003). Yet in classifying such phenomena as mental illness, the DSM makes context “seem not very important or even irrelevant for understanding or alleviating people’s problems” (Boyle 2011, 41). Such a mentality is reflected in the language used to describe such problems in the DSM (Boyle 2011). For instance, Pilgrim and Bentall note that while Brown, Harris, and Hepworth (1995) describe partner violence as a “depressogenic” effect, “this could be reframed by simply stating that miserable women live with oppressive men” (Bentall and Pilgrim 1999, 270).

The Danger of Inappropriate Diagnosis

Another worry with regard to PMDD concerns the potential consequences of such a diagnosis for women. For example, there have been several court cases in which PMS was used as a defence or in sentencing, mainly in the United States, Canada, and Australia in the 1980s (Easteal 1991; Caplan 2004). In In re H, a British case in 1983, a woman was denied custody of her child because it was alleged that she suffered from PMS (D’Emilio 1985), and in a U.S. case, a dentist was acquitted of rape and sodomy because it was alleged that the plaintiff suffered from “premenstrual irrationality” when she reported it (Herbert 1982).

Using severe PMS as a defence is no better, either. For instance, it has been used successfully to mitigate responsibility in certain U.K. murder cases (reducing murder to “manslaughter with diminished responsibility” [Richardson 1995, 765]). Yet such a defence supports the idea that hormonal fluctuations can transform women into violent criminals (Raitt and Zeedyk 2002). It is worth noting that, at least in Canada, before the PMS defence became fashionable in the 1980s, menopause or postnatal depression were more frequently used as defences for minor criminal offences (Easteal 1991). For more serious cases such as infanticide, there lingered an unsubstantiated theory that women who suffer a traumatic childbirth can become insane and are thus not as responsible for their actions as men would be if they kill their child (Sommer 1984).

Going back further, PMS was used to claim temporary insanity in 19th-century England, successfully acquitting one women of shoplifting and two women of murder (Spitz 1987). Whether it be a diagnosis of hysteria,Footnote 2 which goes back thousands of years, or the arguably newer version of hysteria—PMDD—this drive to paint women as being at the mercy of their uterus and “raging hormones” is a longstanding one and reflects an attitude that women are supposed to be gentle and passive by nature, and that to be otherwise is sick and unnatural (Easteal 1991).

The majority of lawyers commenting on the use of PMS as evidence in court have recommended that it be limited to mitigating sentences or reducing charges, placing it in the same category as a history of abuse, recent illness, or lack of a criminal record (Easteal Easteal 1991). Their recommendation is based primarily on the “lack of universal acceptance of PMS as a diagnostic entity; its failure to meet all criteria for an insanity plea; the fear of abuse as a defence by women offenders and the fear of sexism” (Easteal 1991, 5). Now that PMDD is an accepted diagnosis in the DSM, and given that it is thought to be more severe than PMS, the first two impediments to its use in court may have significantly diminished. As Raitt and Zeedyk state, the “existence of a diagnostic category within the DSM communicates the message that it has become generally accepted amongst experts that menstrual functioning is (at least sometimes) the cause of pathological behaviour” (2002, 119). This kind of defence transforms women into entities who may, once a month, be so at the mercy of their hormones that they are rendered temporarily insane and not responsible for their actions, and it reinforces the view that they cannot be trusted in positions of responsibility (Easteal 1991).

Another danger of labelling valid and legitimate anger/distress as PMS or PMDD is that it can shift attention away from the actual cause of that anger/distress. Instead of blaming the abuser and labelling him/her with a problem that should be corrected, a diagnosis of PMDD blames the victim, labelling her with a mental illness that should be corrected. Studies of women’s experiences of “PMS,” such as that by Mooney-Somers and colleagues have found that:

the naming of women’s behavior as “PMS” threatened to negate the individual woman’s experience, such that alternative meanings of her emotion or behavior—meanings that may be uncomfortable or undesirable for her partner, or require relationship work—disappear. Some women described this as resulting in a denial of the negative emotion they were feeling, or the issue they were raising. We also saw that some women experienced their partner naming “PMS” as allowing their partner to deny any responsibility for an issue that emerged premenstrually, because it was just “PMS” (Mooney-Somers, Perz, and Ussher 2008, 73, emphasis original).

(These findings with respect to PMS are consistent with findings for PMDD [Ussher 1996, 2003, 2008].) Thus, pathologising these women’s anger/distress risks adding another layer to their suffering. They would not only suffer from the underlying social issue that is causing their anger/distress, but also from the label of a mental illness that locates the problem within the individual, along with the associated stigma and discrimination and their follow-on consequences. For instance, individuals diagnosed with a mental illness are less likely to have their physical complaints taken seriously (Byrne 1997; Schulze and Angermeyer 2003; Jones, Howard, and Thornicroft 2008). Given that more women are diagnosed with mental illness than men (American Psychiatric Association 2000), this also makes women more susceptible to health disparities.

The Danger of Inappropriate Treatment

If the anger/distress these women are suffering is likely due to stressful life circumstances, then treating their problem as a medical one actually undermines their chances of properly solving it. Providing these women with medication in order to enable them to tolerate an unfair or even harmful situation risks keeping them in conditions that could leave them worse off in the long-term. The medical “solution” then appears akin to the use of alcohol and illicit drugs among some women in stressful situations as self-medication in order to tolerate otherwise intolerable situations (e.g., see the discussion of alcohol and drug use among sex workers in Li, Li, and Stanton 2010).

Relatedly, an alternative to the medical model of recovery for mental illness has emerged called the “place-train” model. While the medical model seeks to train those diagnosed with mental illness to manage their condition before giving them work and accommodation (the “train-place” model), the “place-train” model instead quickly places them in work and accommodation, then gives them training and support (Corrigan and McCracken 2005). Most studies of the “place-train” model have so far shown greater success in helping those diagnosed with mental illness to live independently (e.g., with greater housing stability and quality of life), as well as greater reduction in rehospitalisation rates (Corrigan and McCracken 2005). Thus, even if the issue here is a mental illness, these findings challenge the medical model’s emphasis on medication as the first and most important aspect of recovery.

If we resist the urge to consider any form of distress as a problem within the individual, we could instead look behind the distress to what is most likely causing it as the problem. Just as someone who breaks her leg experiences pain and distress, giving her painkillers might appear to help her by alleviating her pain. However, if the broken leg is not fixed, she may continue to walk on it, oblivious to the further damage she is causing by doing so. The pain associated with the broken leg acts as the body’s alarm system to alert the individual that something is wrong and needs to be addressed.

Likewise, medicating a woman for her premenstrual anger/distress without addressing the problems that lead to it in the first place risks doing her more long-term damage. Given the amount of evidence associating premenstrual anger/distress with stressful life circumstances, presupposing the underlying problem to be a biological one seems misguided. For many women with premenstrual anger/distress (as well as other psychiatric diagnoses such as depression), medical treatment may give the individual immediate relief but may ultimately be a band-aid approach that alleviates the individual’s “symptoms” in the short-term—necessary as this may be—which, without additional or alternative interventions, could still leave her worse off in the long-term.

PMDD Is Culture-Bound

Several studies have found that the mood “symptoms” of PMS and PMDD appear to be suffered more by women in Western countries than in developing countries (Chrisler and Caplan 2002). For instance, in countries such as China and Hong Kong, women report physical premenstrual experiences (such as pain, tiredness, sensitivity to cold) but hardly ever report psychological ones (Chang, Holroyd, and Chau 1995; Yu et al. 1996). Such premenstrual experiences are also accepted in these cultures as normal, not a sign of disease (Epstein 1995). Another study, this time of 1,824 women of different ethnicities in Hilo, Hawai’i, also found that attitudes towards menstruation are influenced by ethnicity (Morrison et al. 2010). In stark contrast to negative attitudes towards menstruation in Western cultures (Delaney, Lupton, and Toth 1988; Luke 1997; Lew et al. 2005; Linton 2007), 60 percent of women in their study described menstruation in positive terms such as “natural” (Morrison et al. 2010). Not only this, but the more time that women of ethnic minorities spend living in the United States, the more likely they are to report PMDD (Pilver et al. 2011). Thus, if we are to accept PMDD as a reified medical disorder, then we must also accept exposure to U.S. culture as a risk factor for contracting PMDD.

If the proponents of the PMDD diagnosis wish to ignore the fact that correlation does not equal causation in their reification of PMDD, then they cannot be selective in their application of that logic (or lack thereof). Such findings were ignored by the Work Group in preparation for the DSM-5, even within studies they actually cited (Epperson et al. 2012). For example, one of the studies cited by the Work Group in fact reports that the rates of moderate to severe PMS and PMDD in Japanese women were much lower than in their Western counterparts. According to the study’s authors, the results suggest that “race and ethnicity influence the expression of premenstrual symptoms” (Takeda et al. 2006, 209). There thus appears to be much evidence that has been overlooked in favour of viewing PMS and PMDD as constructed, culture-bound phenomena rather than medical disorders.

A Desired Effect From Medication Does Not Imply an Underlying Medical Disorder

Despite continued reference to the serotonin hypothesisFootnote 3 by pharmaceutical companies, scientific evidence has failed to support it (Lacasse and Leo 2005). The fact that premenstrual anger/distress can be reduced by taking SSRIs is thus not itself evidence that there was an underlying biological problem that was fixed by the SSRI. Analogously, while Valium may be an effective sedative for those having trouble sleeping, this does not imply that those individuals suffer from an underlying dopamine deficiency. Furthermore, evidence shows that fluoxetine (an SSRI) also reduces “negative affect” (such as sadness and anxiety) in healthy individuals (Knutson et al. 1998). As Caplan argues:

Certainly if fluoxetine is given to any group of depressed or upset people, some will feel better. But that reveals nothing about the causesof their upset and, in this case [PMDD], it doesn’t provide that the upset is related to menstrual cycle-related changes (Caplan 2004, 61).

One could easily imagine that if fluoxetine were given to an African-American slave “suffering” from drapetomania, his distress might be alleviated and he might return to slavery with a smile on his face.Footnote 4 In support of this argument, while most studies of mood and menstrual complaints investigate women only, one study investigated couples and found that women’s moods were “affected less by the menstrual cycle … than by variation in the social environment” and that the men were equally affected by mood swings (LeFevre et al. 1992, 81).

Another medication approved for treating PMDD is a combined oral contraceptive containing drospirenone and a low level of oestrogen (Lopez, Kaptein, and Helmerhorst 2012). A Cochrane Review found that two of the five trials of this oral contraceptive that it assessed showed positive effects for women diagnosed with PMDD compared with placebo (although the placebo effect was quite large) (Lopez, Kaptein, and Helmerhorst 2012).Footnote 5 Yet, as discussed in the following section, a reduction of premenstrual anger/distress by hormonal treatment does not prove that the cause of the anger/distress was an underlying hormonal imbalance.

Correlation Does Not Equal Causation

Firstly, despite many years of research, no correlation between premenstrual anger/distress and hormone levels has been confirmed (Figueira and Dias 2012). Yet even if a correlation were confirmed, this would not prove that the anger/distress is merely a product of faulty biology and is otherwise not significantly related to, if not conditional upon, social/contextual factors affecting the individual. It is worth remembering here that correlation does not equal causation. Thus, even if premenstrual anger/distress is found to be associated with certain biological markers, such a finding is not currently evidence that the biological component preceded the psychological component and not the other way around. It has long been known that acute psychological stress can result in a rise in cortisol levels and changes in the hypothalamus-pituitary-adrenal axis (Hellhammer, Wüst, and Kudielka 2009; Foley and Kirschbaum 2010). (In fact, cortisol levels are widely used as biological measures of psychological stress [Hellhammer, Wüst, and Kudielka 2009].) The hypothalamus-pituitary-adrenal axis is, in turn, also thought to be associated with stress-related pathologies such as systemic hypertension (Esler et al. 2008) and the metabolic syndrome (Chrousos 2000). The fact that our psychology can influence our biology is now well known and also evidenced by phenomena such as the placebo effect, which is now a standard component of many drug trials throughout the world.

It is not implausible to imagine that abuse may cause a change in one’s biological processes and consequently affect the victim’s physical and mental health. In fact, there is growing evidence suggesting exactly that (e.g., Felitti et al. 1998; Kendall-Tackett 2003; Dallam 2005). A 2003 study specifically of PMDD by Girdler et al. indicated that “a history of prior abuse is associated with alterations in physiological reactivity to subsequent mental stress in women” and that even one incident of traumatic abuse, even in (previously) “healthy” women (apparently free of psychological complaints), could have such a result (Girdler et al. 2003, 849). Resnick et al. (1995) showed similar findings and Girdler et al. (2007) found that all women who had suffered abuse showed changes in their biological stress-responsive systems, whether they were diagnosed with PMDD or among the non-PMDD controls. Another study found that a history of sexual abuse was associated with elevated thyroid-stimulating hormone concentrations, regardless of PMS or PMDD (Bunevicius, Leserman, and Girdler 2012). Given the number of studies that report a significant association between a history of physical and sexual abuse and PMS (Koci and Strickland 2007) or PMDD (Paddison et al. 1990; Golding and Taylor 1996; Golding et al. 2000; Girdler et al. 2003), such findings should be taken seriously.

In such scenarios, it would clearly be wrong to diagnose the person who has suffered abuse with a disorder. It is the abuse that ultimately is the root of the problem and concentrating efforts on biological fixes clearly would not address the larger problem of abuse at the heart of the suffering (Riessman 1983; Edwards 1984; Laws, Hey, and Eagan 1985). A biological “fix,” even if it provided some relief, would here be akin to treating the symptoms and not addressing the cause of those symptoms. On an individual level, attempting to alter the biological process alone without also removing the individual from the abusive environment would not only fail to adequately treat the problem but also enable it to continue.

Secondly, even if an individual has certain biological predispositions that cause her to be more sensitive to external stressors (e.g., stressful life events) than other individuals, that does not in and of itself negate the importance of the stressor or mean that the individual is not angry/distressed for good reason. Consequently, it does not mean that it is the biological disposition that should be addressed and not the stressor.

Thirdly, to say that premenstrual anger/distress should not be pathologised does not mean that medication should never be used for it. Medication should here be viewed in the same way that we view the use of paracetamol for headaches. We understand that headaches can be due to a great many things and that the suffering may be clinically significant but not usually a sign of pathology. We also understand that sometimes we take paracetamol for headaches even when those headaches are caused by stressors that could be better addressed by environmental changes in our lives. The fact that the paracetamol works does not mean that one’s headache was due to a paracetamol deficiency (Coyne 1987).

Finally, an increased sensitivity to external stressors is not necessarily a disadvantage. For instance, some individuals who have been diagnosed with depression feel that it acted as a catalyst for change in their lives (Frank 1991; Schreiber 1996; Ridge and Ziebland 2006). In such individuals, depression serves a valuable role—acting as an indicator that something in the person’s life has gone awry and, if so desired by the individual, should be changed. A similar argument has been made for anxiety in social situations and suffering due to trauma—that such reactions are normal and can even be useful (Tedeschi, Park, and Calhoun 1998). In this light, premenstrual anger/distress could be viewed as having a similar function to physical pain—that is, alerting the individual to something that needs to be addressed. The studies presented here indicate that this may well be the case.

Yet people can be uncomfortable with biological variation and too often label it as a problem that needs to be corrected. This can be seen in the relatively recent phenomenon of female genital cosmetic surgery. Reitsma et al.’s (2011) study found that physicians’ willingness to perform surgical labia minora reduction is influenced by their personal opinion regarding how big they think a labia minora should be. Society regards a vulva with a very small labia minora as being more attractive—a preference that has been influenced by the pornography industry (Poelsma, van Erp, and van Lunsen 2000; Borzekowski and Bayer 2005; Koning et al. 2009). This social preference in turn appears to shape a physician’s perception of the “need” for labia minora surgery. I contend that just as the “ideal vulva” is a notion that discourages acceptance of variations in vulva appearance, so too does the feminine ideal also discourage acceptance of variations in female behaviour.

Even if women who experience premenstrual anger/distress might show an increased sensitivity to cyclical variations in hormone levels (Cunningham et al. 2009), and if this may render them more susceptible to react to psychosocial stressors, that in itself could serve a productive function, alerting the individual to something wrong/harmful that should be addressed. Indeed, such an interpretation has been found in North American and Indian studies, in which a significant number of women state they experienced positive aspects of premenstrual changes when given the chance (Alagna and Hamilton 1986; Stewart 1989; Chaturvedi and Chandra 1990; Lee 2002). For instance, a participant in one study finally decided that there was meaning to be found in the issues that kept resurfacing every month: “I’ve been repressing things for years … just saying, ‘Oh, it’s just PMS.’ That’s bullshit. You need to look at it and say, ‘Well, if I keep feeling this way, about this particular issue, at this time of the month … then, it’s got to mean something’” (Ussher and Perz 2013, 143). Viewed in this light, an increased sensitivity could be considered an advantage that some women have once a month that enables them to be alerted to issues in their lives that need to be addressed. To medicate them would be to dampen potentially productive responses to these issues, keep them in unfair or harmful situations, and present a barrier to progress in their lives. As Warsh succinctly summarises:

By maintaining a disabled, inferior view of female physiology, structural inequalities such as inadequate financial resources, a double or triple daily workload, and a lack of autonomy that also create stress and exhaustion are ignored. In this fashion, the medicalization of menstruation, and the related labelling of womanhood as disability, continues to support gendered socio-economic and political inequities (Warsh 2010 , 37).

This tendency to label variations from the norm as “disorder” and to subsequently “treat” them is of course not particular to debates about PMDD. Intersex peopleFootnote 6 can be labelled as having a disability or a “disorder of sex development” (Community Affairs References Committee 2013, 1).

Such a mentality leads doctors and parents to believe that “treating” this “disability” is appropriate in order for the child to conform to societal expectations of the physical features of each sex. “Treatment” consists of involuntary or coerced sterilisations or genital surgery.

For example, the recent 2013 Australian Senate inquiry on involuntary or coerced sterilisation of people with disabilities presented the case of child A, who was born with a clitoris-phallus (a clitoris that was considered too big) so surgery was undertaken to make the baby look more feminine (Community Affairs References Committee 2013). In a non-intersex girl, this would be considered female genital mutilation, yet because being intersex is considered a form of disability, the surgery is allowed. When child A grew up, she wished to change sex and become a boy. This is not uncommon, as the Senate inquiry reports that 8 percent to 20 percent, and in some cases up to 40 percent, of intersex people have gender issues as adults (Community Affairs References Committee 2013). Also sadly prevalent are the rates of depression, trauma, self-harm, and suicidal tendencies experienced by intersex people who have been through such surgeries and sterilisations. Whether the discussion is about intersex children or PMDD, we can see that the label of “disability” or “disorder” sanctions “treatments” that, if the features in question were to be viewed as simply variations rather than pathologies, would be regarded as harmful and unjust.

That the Anger/Distress Seems to Occur Monthly Is Not Evidence That the Problem Is Linked to the Menstrual Cycle Rather Than the Social Context

To assume otherwise would be to ignore the long history of psychosomatic illness and how our experience and reporting of distress is shaped by culture. Basically, the above assumption ignores how much our states of mental health or illness are socially constructed. As Shorter describes, “patients’ representations of disease help form the contents of the symptom pool. What people thought they had plays an important role in the history of psychosomatic illness, the more so because these attributions are especially vulnerable to medical ‘shaping’” (Shorter 1992, 10; cf. Hacking 1995). As we are inextricable from our culture, so are our experiences and representations of distress. What strengthens this view even more in relation to PMDD is Cosgrove and Riddle’s (Cosgrove and Riddle 2003) finding that women often expect to suffer from premenstrual anger/distress. When understood alongside studies that show that when patients are told something will be painful the patient will experience it as even more painful (e.g., Koyama et al. 2005; Benedetti 2007; Benedetti et al. 2007), we begin to appreciate the significance of cultural expectations of premenstrual anger/distress. Cultural expectations affect the expression of illness and, as recounted earlier, this is no different in PMDD. In this way, an individual’s preconceived notions and culturally-bound expectations of the patterns that women’s moods should take affect what she is alert to, notices, and reports.

Added to this is the “pressure cooker” theory, in which the focus is not on asking why the woman “explodes” with anger/distress once a month but why she represses it for three out of four weeks (Ussher 2003). According to this model, premenstrual anger/distress could be understood as an understandable reaction to (for example) one’s family situation, in which women are typically burdened with much greater responsibilities and are expected to bear them silently and put others’ needs above their own (Jack 1991).

This is where culture enters. According to this alternative explanation for why the anger/distress seems to occur once a month, women are not expected to express such emotions; once a month, however, during the premenstrual phase, such mood and behaviour are allowed and expected (Kendall 1991). By internalising such cultural attitudes, women express their anger/distress during this period because it is the culturally-sanctioned time for it and can be biologically “excused” (Koeske and Koeske 1975; Koeske 1983, 1987; Bains and Slade 1988; McFarlane, Martin, and Williams 1988). Warsh calls it a temporary, socially-sanctioned “inversion of accepted behaviour” akin to the only time in the year when it is acceptable to demand sweets from strangers (Halloween) or the early modern folk tradition of charivaris (Warsh 2010, 36).Footnote 7 For example, Pirie’s study of illness in women found that many of them “could only legitimate feelings of anger and frustration (largely arising out of dual career obligations) when they were understood to be biological rather than social relational responses” (Pirie 1988, 642). Yet women should be able to legitimate anger/distress regardless of how it is understood. In order for them to do so, we must accept distress outside of the constraint of DSM-legitimisation.

Finally, even if certain behaviours were shown to occur cyclically in some individuals due to a biological mechanism, that in itself would not show that the behaviour is pathological. As Raitt and Zeedyk succinctly state: “It is possible to acknowledge the existence of menstrual phenomena without defining those symptoms as illness. … The fact that certain events occur together or recur in a predictable pattern does not necessarily mean that they are linked or that they are an illness” (Raitt and Zeedyk 2002, 119–120). For example, Allen wittily suggests a “Pre-Breakfast Syndrome” that “lumps together all the various complaints which could ever, in any individual, be shown to appear regularly in the first hours after waking and then to subside” (Allen 1984, 30). Symptoms would include “habitual hangover, morning sickness, smoker’s cough, lethargy or excitability, reduced or increased libido, irritability, intellectual impairment, and numerous others” (Allen 1984, 30).

The Discrepancy in Rates of Diagnosis of Mental Disorder Between Men and Women Neither Provides Evidence of a Biological Difference nor the Need for an Additional Diagnosis

According to the authors of the DSM-IV-TR (American Psychiatric Association 2000), women are two to three times more likely than men to be diagnosed with dysthymia (a form of chronic depression), twice as likely to be diagnosed with major depressive disorder, and three times more likely to be diagnosed with panic disorder. Moreover, more than 90 percent of people diagnosed with anorexia and bulimia are women. This is taken as evidence that women are genetically more predisposed to suffer from psychiatric disorders. However, the logic here is faulty, as it is akin to saying that if more men suffer from cancer, it is evidence that mesothelioma (lung cancer associated with asbestos) is caused by a genetic abnormality. The assumption thus inherent in the DSM ignores the social inequalities affecting women (Ali, Caplan, and Fagnant 2010) and the greater likelihood of women to seek professional help compared to men, resulting in more psychiatric diagnoses (Padesky and Hammen 1981; Thom 1986; Husaini, Moore, and Cain 1994; Möller-Leimkühler 2002). In essence, the argument makes the mistake of conflating correlation with causation.

Women experience disproportionate levels of stress and trauma in comparison to men, so it is not surprising that they suffer from disproportionate levels of psychological distress (Ali, Caplan, and Fagnant 2010). In depression, for example, differences between men’s and women’s lives that make the latter more at risk of depression are by now well documented (Astbury and Cabral 2000; Belle and Doucet 2003; Stoppard and McMullen 2003). The approach that focuses on women’s lives explains depression in women as primarily a result of the gendered power imbalance in a patriarchal society (Stoppard 2000).

Patriarchal society also impacts men. For instance, it has been reported that men are less likely to report to the doctor for signs of depression (Rogers, May, and Oliver 2001) due to the pressure to conform to a masculine image that discourages disclosure (Rogers and Pilgrim 2009), resulting in fewer diagnoses. In societies with prevalent sectors that emphasise control and strength as masculine attributes, any weakness or loss of control perceived as inherent in seeking help for emotional distress would discourage men from seeing the doctor (Emslie et al. 2006). Another factor that accounts for the increased prevalence of the diagnosis of depression among women is that women have longer life expectancies than men, which means there are more women still surviving in old age to be affected and that they are more likely to have lost a partner. Women also tend to have lower-paying jobs than men, and low-paid work is linked with worse mental health (Rogers and Pilgrim 2009).

In summary, there are several sociological explanations for the differing rates of diagnosis of affective disorders between the genders, whereas a biological explanation for this disparity has not emerged. Even if one were to emerge, it would not prove that biological factors are the major, let alone sole, explanation for gender differences in rates of affective disorders.

Importantly, even if the inequality in psychiatric diagnoses between men and women were due to genetic differences, this does not justify the need for an additional diagnosis. Evidence must be presented for the existence of a distinct disorder each time a new one is proposed, but arguing that women are more likely to suffer from psychiatric disorders anyway does not explain why additional disorders specifically for women should be added to the DSM rather than working with existing diagnoses.

The Risk of Over-Diagnosis

The Psychological Slippery Slope

While it may be inappropriate to stigmatise women as a group based on the experience of a subset, recognising this in no way addresses the risk of it occurring. This risk can be described as the psychological (or empirical) form of the slippery slope. That is, the psychological or social effects of accepting A mean we will eventually accept B. This could occur if allowing A influences a change in our intuitions such that our moral opinions about B also change (Van der Burg 1991). In the case of PMDD, accepting that some women are subject to a mental disorder that interferes with their responsibilities could result in a social psyche that renders any woman potentially suspect of this disorder. Belief that women are subject to raging hormones that interfere with their behaviour and decrease their intellectual capacity (Dalton 1979) has already been used to argue that women are not fit for positions of responsibility. For example, in the 1970s, Australian commercial airlines used this idea to argue that women were unfit to become pilots (Kaplan and Rogers 1990). Edgar Berman, a prominent U.S. physician, used the same idea in 1968 to argue against having female physicians and presidents (Fausto-Sterling 1985). Furthermore, PMDD may itself be understood as the product of a psychological slippery slope in psychiatry—an extension of an older diagnosis, hysteria, as discussed earlier.

Risk of Over-Diagnosis Via GP Prescribing

That this “condition” supposedly occurs in a minority of women also does nothing to mitigate the risk of over-diagnosis. One path to over-diagnosis may come from an increased awareness of PMDD among general practitioners (GPs) due to its move to a full category in the DSM.Footnote 8 As the Work Group states, the move aims to boost the legitimacy of PMDD as a medical disorder (Epperson et al. 2012), which will encourage GPs to apply the diagnosis (and similar concerns have been voiced with regard to other mental illnesses). Yet it would not be difficult to conflate PMS with PMDD. With wide-ranging “symptoms,” PMS is said to affect between 30 to 40 percent (Pearlstein 2007) and 75 percent (Steiner 2000) of menstruating women. Hence, an increasing number of women could be labelled with a mental disorder that renders them angry, distressed, or depressed once a month.

Risk of Over-Diagnosis Via Patient Self-Diagnosis

Given that awareness of PMDD is likely to rise with its inclusion as a full category in the DSM-5, this increases the likelihood that more women will judge themselves to have PMDD and present to their GP with their self-diagnosis. This is especially likely in countries such as the United States and New Zealand that allow direct-to-consumer advertising, as research has shown that women are more likely to pathologise any premenstrual negative moods and experiences when they are introduced to PMDD as a medical disorder (Nash and Chrisler 1997). Even the occasional concurrence of a negative mood with the premenstrual phase is enough to maintain the belief in PMDD (McFarlane, Martin, and Williams 1988). There is also the risk that women will conflate PMS and PMDD. Advertisements for Sarafem already conflate the two (Caplan 2008). Encouraging such a conflation is a way of expanding the number of women who could be diagnosed with PMDD, thereby increasing sales of Sarafem.

History Suggests Risk of Over-Diagnosis Is Real

If we look at the recent history of bipolar disorder, we find an example of the above arguments in relation to PMDD. Moreno et al. (2007) reported a 40-fold increase in the diagnosis of bipolar disorder in children and adolescents in the United States between 1994 and 2003. This increase was not driven by DSM criteria but appears to have come from a change in diagnostic practice among clinicians (Mitchell, Loo, and Gould 2010).

The steep rise in diagnosis of childhood bipolar disorder appears to have fed off a number of messages, including those that suggested that bipolar disorder in children is greatly under-diagnosed (Geller, Fox, and Clark 1994; Biederman et al. 1995; Geller et al. 1995; Wozniak et al. 1995) and has more diverse presenting symptoms (including irritability and emotional liability) (Carlson 1983). A parallel with PMDD lies in the potential to conflate PMS with PMDD that could lead to the belief that PMDD is in fact extremely common and was previously under-diagnosed. Moreover, the leap from PMS to PMDD is not as large as the leap from the usual criteria for bipolar disorder to how clinicians believed bipolar disorder presents in children.

Depression provides another example of this phenomenon of over-diagnosis. When imipramine, the first antidepressant, was developed, it was not expected to sell well because it was thought there were not many people with depression (Healy 1997). Compare that with the current situation, in which 11 percent of people over the age of 11 in the United States are currently taking an antidepressant (Pratt, Brody, and Gu 2011). For U.S. women in their 40s and 50s, 23 percent take antidepressants (Pratt, Brody, and Gu 2011). Of course, as with bipolar disorder, it could be argued that the huge rise in diagnosis is because depression was previously under-diagnosed. However, a meta-analysis of 41 studies, capturing 50,371 patients, found that for every 100 cases in primary care, 15 were false positives compared to 10 false negatives and 10 identified cases (Mitchell, Vaze, and Rao 2009). Moreover, this figure is judged by standard diagnostic criteria—criteria that were loose to begin with (Parker 2007) and are now even looser in the DSM-5 with the removal of the bereavement exclusion (American Psychiatric Association 2013a).

Two important factors have contributed to the steep rise in depression diagnoses. Firstly, the classification of depression changed drastically with the introduction of the DSM-III (Parker 2007). Prior to the DSM-III, depression was classified as either reactive/exogenous (triggered by stressful life events) or endogenous (unrelated to external stressors) (Parker 2007). In the DSM-III, this distinction was removed and replaced with the term “major depression” and criteria that were set “at the lowest order of inference” (Parker 2007). The change in classification came in conjunction with a change in mindset—in depression, the patient’s context was pushed aside (except in the case of bereavement) and biological correlates instead took centre stage in the discourse surrounding the cause and treatment of depression (Callahan and Berrios 2005). Secondly, the rate of diagnosis of depression rose sharply when the first SSRIs were introduced (Healy 2004). Their introduction was accompanied by heavy marketing, both to clinicians and to consumers in the United States (Gøtzsche 2013), and went hand-in-hand with the focus on biological explanations of depression.

Given the similarities between the histories of bipolar disorder and depression diagnoses on one hand and the evolution of the PMDD diagnosis on the other, there is cause to take the risk of over-diagnosis of PMDD seriously.

Policy Suggestion

I have argued that for women who present with premenstrual anger/distress, it is unjustified as well as unethical to describe their anger/distress as a mental disorder. Not only has the biological “evidence” for PMDD failed to substantiate it as a mental disorder, but doing so also: (a) is likely to inaccurately locate the problem exclusively within the individual rather than also in her life situation, (b) results in stigma, which translates to individuals labelled as mentally ill not being taken seriously when they suffer physical symptoms, and (c) may undermine attempts to address the root of the problem. This is because medicating these women enables them to tolerate harmful situations and risks keeping them in conditions that leave them worse off in the long-term. Alleviation of distress by medication may thus be outweighed by long-term injustices and harms to both the women suffering them and women in general, and not only in the form of stigma. There are therefore strong reasons to recommend that the PMDD diagnosis be removed from the DSM altogether. PMDD also should not be included in the upcoming ICD-11, as suggested by the beta draft of the ICD-11 (World Health Organization 2013). (In the current edition, premenstrual tension syndrome, rather than PMDD, is listed. The threshold for diagnosis is lower than for PMDD and it is not categorised as a mental illness but under “Pain and other conditions associated with female genital organs and menstrual cycle” [World Health Organization 2010].)

Having said this, I do not wish to say that it is not possible for a woman to experience premenstrual anger/distress that is totally unrelated to any life stress. There may well be a small percentage of women who experience premenstrual anger/distress that cannot be explained or significantly associated with any life event or environmental stress. I suggest that such women should be able to receive treatment without being diagnosed as having a mental disorder. This approach is already taken for women who experience very heavy, painful periods and are prescribed the contraceptive pill and for individuals who experience headaches and take paracetamol. A diagnosis of “painful period syndrome” or “painful head syndrome” is not required for such individuals, nor should it be required for women experiencing premenstrual anger/distress unrelated to life stress.Footnote 9

Conclusion

The evidence presented in this paper points towards PMDD as a socially constructed diagnosis that could have detrimental effects for women. The main concern of the feminist argument is with pathologising the anger/distress that women have good reason to feel and that are indicative of external issues in their lives that should be addressed (Ussher 2011). The feminist argument against the PMDD diagnosis further states that women should not need to be labelled as mentally ill in order for their anger/distress to be taken seriously (Caplan 2004). It is thus inappropriate to categorise and label women’s premenstrual anger/distress as a mental disorder. Experiences can be of clinical significance, but this cannot be misconstrued as sufficient grounds for a determination of pathology. I therefore suggest removing PMDD from the DSM and the ICD. There also should be more research on the social issues that both feed into and spring from the diagnosis of PMDD, and the existing sociological research on the matter should be taken more seriously by psychiatry, especially in relation to psychiatric classification.