Dyspareunia in postmenopausal women: A critical review

1Department of Psychology, McGill University; 2Department of Psychology, Sex and Couple Therapy Service, McGill University Health Centre (Royal Victoria Hospital); 3Department of Obstetrics and Gynecology, Sir Mortimer B Davis Jewish General Hospital, Montreal, Quebec Correspondence: Ms Alina Kao, Department of Psychology, McGill University, 1205 Docteur Penfield Avenue, Montreal, Quebec H3A 1B1. Telephone 514-398-5323, fax 514-398-4896, email akao@ego.psych.mcgill.ca A Kao, YM Binik, A Kapuscinski, S Khalifé. Dyspareunia in postmenopausal women: A critical review. Pain Res Manage 2008;13(3):243-254.

As with most female sexual difficulties occurring during midlife and beyond, dyspareunia is typically considered to be a consequence of declining ovarian hormone levels (6,7). As a result of aging tissue and decreasing levels of endogenously produced estrogens during menopause, in particular estradiol (E 2 ), atrophic changes may be observed in the external genital region, introitus and vagina (ie, vaginal atrophy). The resulting symptoms may include itching, vulvar pruritus, vaginal dryness and dyspareunia (8). Not surprisingly, medical and pharmaceutical attention has focused on postmenopausal dyspareunia and its relationship to estrogens. Consequently, hormone replacement therapy (HRT) has long been considered the frontline and almost exclusive treatment for painful intercourse in postmenopausal women (7,9). Considering the many other factors that may negatively affect sexual functioning in postmenopausal women, such as increased likelihood of health problems, pelvic floor muscular dysfunction, dermatological conditions, relationship factors, partner's health and psychosocial stressors relevant to midlife (eg, role changes within the family, caring for elderly parents, retirement), this seems to be an oversimplified and one-dimensional view of dyspareunia in postmenopausal women. Moreover, vulvar and pelvic pain conditions commonly affecting premenopausal women (eg, provoked vestibulodynia, vulvodynia, vaginismus) may also exist in postmenopausal women and have been more or less ignored. This seems odd because the main presenting complaint is pain. The lack of interest in the actual pain is reflected by the almost total absence of clinical or research reports about postmenopausal dyspareunia in the pain literature. The present review examines the traditional and widely held conceptualization of postmenopausal dyspareunia as a relatively direct symptom of hormonal decline.
To identify research relevant to the present review, searches of the electronic databases PsycInfo, MEDLINE, Biosis and Excerpta Medica were performed for articles and empirical studies published in English-language journals between January 1969 and August 2007 concerning the prevalence, etiology or treatment of dyspareunia in perimenopausal or postmenopausal women. Hence, the keywords dyspareunia, coital pain, sexual function, sexuality, sexual dysfunction, vaginitis, urogenital, genital or vaginal were combined with the keywords menopause, climacteric, perimenopause, postmenopause, aging, hormone, estrogen or hormone therapy. In addition, a manual search of pertinent books, journals and a review of the reference sections of included papers was performed. The first author performed database searches, reviewed germane articles to extract research findings and evaluated study quality by the type of research protocol used. The coauthors were involved in the research evaluation and verification process.

GENERAL METHODOLOGICAL ISSUES
Before discussing the available evidence, it is important to consider general methodological issues that were commonly encountered in the reviewed articles. In critically assessing the past literature, differences in study design and subject variables made it difficult to compare or aggregate findings from different investigations. For example, a widespread shortcoming of the reviewed literature was the failure to report or distinguish between reproductive stages (eg, [10][11][12][13][14][15]. Age alone, and not menstrual status, was often used to gauge the reproductive phases of participants. This is problematic, because there is a large variance in the age at which individual women experience menopause. As such, the effects of age and menopause were confounding variables in these investigations. Moreover, discrepancies existed in how reproductive phases were defined. The World Health Organization (16) suggested the standardized definitions for the stages of the female reproductive life cycle listed in Table 1. These World Health Organization definitions will be used in the present review unless otherwise specified.
Another common methodological limitation of the literature was the failure to indicate whether women were on HRT (eg, 4,17-21). If HRT was specified, studies often failed to distinguish between the varying replacement regimens which can have potentially different effects on the urogenital system and sexual functioning (eg, [22][23][24]. Studies often also failed to differentiate between surgically and naturally menopausal women (eg, 20,25-27). This is an important factor to consider because hysterectomy and oophorectomy may cause altered sexual response and functioning (28)(29)(30).
The failure to use standardized measures of pain and sexuality was another frequent limitation of the literature (eg, 17,22,31,32). Standardized and validated pain measures, such as the McGill-Melzack Pain Questionnaire (33), were never used. Also, although there are few standardized and validated assessment instruments of female sexual dysfunction for use in postmenopausal women, such as the McCoy Female Sexuality Questionnaire (34) and the Female Sexual Function Index (35), their use would have been valuable in disentangling the occurrence of different problems. For example, the symptoms of lack of lubrication during intercourse and pain during intercourse were often grouped together (eg, 20,36-39). These problems can occur separately and may have different causes.

PREVALENCE
Population-based prevalence studies of dyspareunia were heterogeneous in research methodologies and sample populations. Many studies included women who were likely to be postmenopausal because they were middle-aged or older. They could not be integrated into the present summary because they did not report or separately analyze findings by reproductive phase (eg, [10][11][12][13][14][15]40). Studies were also excluded if they clustered other symptoms, such as lack of lubrication or vaginal dryness, together with dyspareunia (eg, 36,37).
As Table 2 illustrates, findings from population-based studies are disparate. Estimates of point prevalence of postmenopausal dyspareunia derived from large-scale studies range from 2% (22) to 29% (41). Recent studies (eg, 24,41) have obtained higher prevalence estimates than older investigations (eg, 17,18), which may be, in part, a product of changing attitudes about sexual functioning in aging populations. For example, some studies only queried sexually active women about sexual difficulties or pain during intercourse (eg, 18,42). As such, they may have under-represented their findings, because many women do not engage in sexual activity because it is too painful. On the other hand, the differences may, at least in part, reflect sociocultural, lifestyle, biological and variations of reporting methods among differing populations. For example, major international differences in rates of dyspareunia in 40-to 80-year-old women have been found in a global investigation of 29 countries (15). The Northern European countries had the lowest point prevalence (5%), while the Southeast Asian countries had the highest (22%). Moreover, ethnic differences in reported rates of dyspareunia in premenopausal and perimenopausal women have also been found (42). The most recent population-based prevalence investigations specifically examining postmenopausal women not taking HRT estimated that between 21.5% and 29% suffer from dyspareunia (24,41). Studies with large sample sizes that included HRT users and nonusers obtained lower prevalence estimates of postmenopausal dyspareunia, ranging from 2% to 6.7% (22,23). Patient population prevalence studies are important because they help to elucidate the clinical load of a particular problem in the health care setting. Similar to population-based studies, Table 3 demonstrates that findings on the prevalence Age presented as mean ± SD and/or range. All studies were cross-sectional. Recruitment: a Random selection (eg, from a census), b Representative sample of a specific population, c Volunteer (eg, from advertisements), d Convenience sample (eg, subsample from an ongoing study); Reproductive phases: 1 Premenopausal, 2 Perimenopausal, 3 Postmenopausal, 3a Early postmenopausal, 3b Late postmenopausal, 3c Naturally postmenopausal specified, 3d Surgically postmenopausal specified; Measures of dyspareunia: 2 Adhoc structured questionnaire or interview; Administration: a Face-to-face, b Telephone, c Postal. HRT Hormone replacement therapy of postmenopausal dyspareunia in outpatient samples were discrepant across investigations; they range from 11% (25) to 45.3% (43), and generally fell between 35% and 45%. The reported range of clinical outpatient rates was remarkably greater than the range stemming from nonclinical samples (ie, 2% to 29%) (23,41). One reason for this difference in reporting may be that women may perceive dyspareunia as a 'valid' medical problem when it is being queried in the context of a health care setting. As a result, they may feel less embarrassed about disclosing pain experienced during sexual intercourse. Alternatively, dyspareunia may be one of the reasons women seek treatment in special clinics. Because of this possibility, it is important to remark that a significant proportion of treatmentseeking postmenopausal women report dyspareunia when asked.

CLINICAL PRESENTATION
Subtypes of premenopausal dyspareunia can be categorized by differences in location, intensity, temporal pattern and sensory quality (44). These variables are similar to those used in the pain classification system advocated by the International Association for the Study of Pain (45) as well as the International Society for the Study of Vulvovaginal Disease (46) and are considered useful markers for different etiologies (47). However, very limited data is available on the clinical description of dyspareunia affecting postmenopausal women. By and large, investigations have failed to assess the anatomical location of pain experienced during intercourse and have not enquired about specific pain characteristics in postmenopausal women. Two somewhat vague anatomical locations of dyspareunic pain, 'superficial' (ie, external genital) and 'deep' (ie, internal genital or pelvic), have been assessed in postmenopausal women. Superficial provoked pain on contact, particularly in the vulvar area, is the most common form of premenopausal dyspareunia, affecting an estimated 12% of women (48). From the meagre data available, superficial dyspareunia also seems be the most frequently occurring subtype later in life, affecting an estimated 8.5% of women 50 years of age or older (reproductive phase not reported) in the general population (48). The preponderance of superficial pain is also found in clinical populations, and is estimated to affect 22% of clinical samples of perimenopausal and postmenopausal women (27), and 32% of gynecological outpatients 50 years of age or older (12). However, whether the preponderance of such cases includes vulvar pain is unknown. Although there is a paucity of information about deep dyspareunia in the general population of postmenopausal women, it has been found to be less prevalent than superficial dyspareunia in clinical samples. Estimated rates of deep pain during intercourse range between 8% and 16% in outpatient samples (27,12).
Intensity of postmenopausal dyspareunic pain was assessed in a sole study examining a group of gynecological outpatients and volunteers; it was found that the majority of sufferers, 37.4% of women living with their spouses, experienced slight pain or discomfort, while 6.1% of women living with their spouses had moderate to severe pain (21). At the time of writing, only one study has investigated the sensory quality of dyspareunic pain in middle-aged women; although feelings of either sharp contact pain, or itching or burning sensations were endorsed, there was no enquiry about other types of pain (eg, aching, pinching, stretching) (48). Even though data concerning the clinical presentation of postmenopausal dyspareunia are scarce, it would seem from the information available that there is variation in the location and sensory quality of pain (12,48,49). It appears that the most common subtype is superficial provoked pain that has a sharp or burning quality, which is remarkably similar to provoked vestibulodynia (also known as vulvar vestibulitis syndrome), the most common form of dyspareunic pain in premenopausal women.

RISK FACTORS AND ETIOLOGY
In the past decade, the literature concerning the causes of premenopausal dyspareunia has rapidly grown. Numerous etiological and maintaining mechanisms, both organic and psychological, have been proposed (eg, history of recurrent yeast infections, elevated genetic susceptibility to inflammatory disorders, heightened anxiety and stress) (47,(50)(51)(52). However, this growth in the premenopausal etiological literature has not carried over to postmenopausal dyspareunia. Comparable dyspareunic pain occurring during or after the menopausal transition has almost unanimously been attributed to aging, decreased estrogen levels in the genital tract, and resulting vaginal dryness and atrophy (53). Related to these changes, decreased sexual arousal and lack of lubrication are other proposed mechanisms responsible for pain during intercourse (7).

Age
The process of aging involves not only hormonal alterations, but also other physiological, psychological and social changes that can compromise women's sexual functioning and activities (54). Although the findings have been mixed, the majority of large-scale cross-sectional studies (ie, n>1000) report that the occurrence of dyspareunia decreases with age. Two crosssectional studies of women between the ages of 18 and 59 years found that reports of dyspareunia decreased with increasing age (n=1622 [13]; n=908 [14]). However, this seems to mirror a decline in the number of women engaging in sexual activity with progressing age in the American sample (55). Parallel declines in pain and overall sexual activity were also found with age in a British cohort of postmenopausal women (n=2045) (22). A pan-European study with a sample of 3062 women between 55 and 75 years of age reported significantly lower rates of dyspareunia in older women (23). On the other hand, two cross-sectional studies that included wide participant age ranges (ie, 12 to 78 years old, n=887; 35 to 59 years old, n=436) found that the frequency of dyspareunia increased with the age of the population (11,12). A third cross-sectional and longitudinal study (n=201) also presented findings of increased prevalence of combined dyspareunia and vaginal dryness with age (39), although whether dyspareunia specifically increased was not reported. Another study found that the occurrence of dyspareunic pain fluctuated with increasing age (n=1761) (40). Although women 60 to 65 years of age reported more dyspareunia than those between 50 and 55 years of age, the rates decreased and stabilized for women older than 65 years. However, Weber et al (n=104) (56) found no age differences between outpatient women who reported dyspareunia, vaginal dryness or both, and women who did not have either symptom. Evidence from the majority of large-scale cross-sectional investigations suggests that increasing age may be associated with decreased rates of dyspareunia (13,22,23), but this may be ascribed in part to declining sexual activity in older women.

Reproductive phase
The influence of the menopausal transition is often confounded by aging if these variables are not statistically controlled for in studies of midlife sexual functioning. Menopausal status, independent of the effects of aging, has been associated with a reported "decline in sex life" and difficulties with intercourse (57). There is some evidence in the reviewed literature that suggests part of these difficulties involve pain and that women experience dyspareunia with augmented frequency after menopause. For example, the prospective Melbourne Women's Midlife Health Project found that dyspareunia became significantly more frequent as women went from the premenopausal to the postmenopausal phase (58,59). Moreover, Moore and Kombe (38) found a greater occurrence of dyspareunia in women in late menopause (ie, six to 10 years postmenopause) compared with early menopausal women. A cross-sectional investigation showed that superficial, not deep, dyspareunic pain increased with menopausal status (27).
On the other hand, findings from several other investigations suggest that menopause may not be the main risk factor for the development of dyspareunia. For example, Rekers et al (18) found that despite an increased rate of dyspareunia in sexually active postmenopausal women, the overall rate of vaginal pain was similar for women in the premenopausal and postmenopausal phases. By separately analyzing women in different reproductive stages, Rosen et al (26) found that pain during intercourse affected 24.5% of premenopausal women, 24.6% of perimenopausal women and 34.7% of postmenopausal women. However, the increased rate of dyspareunia in postmenopausal women was not large enough to be statistically significant. Furthermore, no direct relationship was found between later reproductive phases and dyspareunia in three other investigations (17,56,60). Although there is some evidence from the reviewed research suggesting that women do experience coital pain more frequently after menopause, there are a comparable number of contrary findings reported. This makes it difficult to conclude whether dyspareunia is affected by reproductive stage and calls into question the existence of a specifically postmenopausal dyspareunic syndrome.

Decreasing hormones and vaginal atrophy
Although decreasing levels of endogenously produced hormones are the hallmark of the menopausal transition and are thought to be directly linked to postmenopausal dyspareunic pain, only five studies were found that actually examine their association (59)(60)(61)(62)(63). Table 4 summarizes findings from research on the relationship among dyspareunia, hormone levels and vaginal atrophy in postmenopausal women. Vaginal atrophy is a manifestation of aging tissue, and the cytological and chemical transformations in the genital region that result from declining levels of estrogens, particularly E 2 , during menopause. It is purported to be the primary cause of postmenopausal dyspareunia (53). Furthermore, lack of estrogen can also lead to vaginal narrowing and shortening, thereby increasing the likelihood of painful intercourse (64). Moreover, the drop in circulating estrogens is thought to be the cause of vascular changes resulting in diminished physiological arousal to sexual stimulation and lack of lubrication, which consequentially lead to dyspareunia (7).
In examining eight years of longitudinal data from the Melbourne Women's Midlife Health Project, Dennerstein et al (59) used structural equation modelling to demonstrate that declining E 2 affects dyspareunia. This model predicts that increasing E 2 levels equivalent to those in premenopausal women at midcycle would be necessary to decrease dyspareunia. On the other hand, the previous level of dyspareunia was also a major determining factor that predicted current coital pain. This was the only study found in the postmenopausal literature to examine pre-existing dyspareunia. Although the duration of previous levels of dyspareunia was not specified, the results suggest that many postmenopausal women in this sample did not develop dyspareunia solely due to declining estrogen. It is possible that a proportion of postmenopausal women developed dyspareunia before menopause. Another study (60) reported that menopausal status and lowered E 2 were associated with dyspareunia. In contrast, three other studies found no direct relationship between estrogen levels and reports of dyspareunia (61)(62)(63); it is possible that due to their smaller sample sizes, these studies lacked sufficient statistical power to detect subtle effects of E 2 . No evidence of association was found between dyspareunic pain and levels of testosterone, androgens, follicle stimulating hormone or luteinizing hormone (61,63), which are hormones generally associated with sexual functioning.
The occurrence of dyspareunic pain in postmenopausal women has, by and large, been ascribed to tissue aging and the progressively atrophizing effects of declining estrogens within the female genitalia. Yet, in examining the reviewed articles (Table 4), not one study concluded that dyspareunia in postmenopausal women was associated with either occurrence or severity of vaginal atrophy (56,61,63).
Although lower E 2 levels may contribute to the development of dyspareunia in postmenopausal women, a direct association has not consistently been found. Vaginal atrophy, a sequela of the lack of estrogens in the genital region, has also not been found to be associated with dyspareunia. These findings suggest that other nonhormonal causal factors may be involved in the development of dyspareunia in postmenopausal women.

Decreased arousal and lack of lubrication
Decreased physiological arousal and lack of lubrication during intercourse are frequently reported sexual difficulties in menopause and are commonly attributed to decreasing estrogen levels (7). Certainly, lack of lubrication increases friction during intercourse and can lead to pain (7,39). Although inadequate lubrication has been thought by some to be the chief cause of postmenopausal dyspareunic pain, the prevalence of lubrication difficulties has been found to increase with age (14) or remain stable (13), while the majority of large-scale studies found that dyspareunia decreased with age (13,14,22,23). Moreover, a psychophysiological investigation using vaginal photoplethysmography demonstrated that although baseline vasculogenic differences exist between postmenopausal women who reported moderate dyspareunia and premenopausal women, both groups showed similar genital responsivity to visual erotic stimuli (63). The available evidence suggests that even though lack of lubrication and physiological arousal may cause coital pain, their association with dyspareunia in postmenopausal women is weak.

Pelvic floor abnormalities
Pelvic floor abnormalities have been associated with menopause (65,66). The pelvic floor musculature can be pathologically dys-synergistic, hypertonic or hypotonic (67). Postmenopausal women are more commonly affected by pelvic floor hypotonus (65,66,68). This may cause deep dyspareunia due to lack of pelvic stability (67). Furthermore, pelvic floor functioning abnormalities may develop consequent to an initial episode of dyspareunia triggered by a pathophysiological cause (eg, infection, vaginal atrophy, vestibulodynia). The pelvic floor dysfunction may persist after the original cause of dyspareunia has been treated and may become a maintaining factor of coital pain (69). At the time of writing, there are no studies that assessed the effect of pelvic floor functioning on dyspareunic pain in postmenopausal women.
Available etiological evidence suggests that some postmenopausal women may develop pain during intercourse associated with their menopausal transition status, impoverished E 2 levels, decreased sexual arousal or lack of lubrication. However, the inconsistent nature of the reviewed findings suggests that other mechanisms may be involved. It is likely that, similar to premenopausal dyspareunia, multiple etiologies exist for postmenopausal dyspareunia and that these vary among women.

TREATMENT
HRT is widely considered to be the primary intervention for postmenopausal dyspareunia (70). It is suggested that women wishing to avoid the use of HRT should be offered nonhormonal vaginal moisturizer (ie, Replens, WellSpring Pharmaceutical, Canada) (70). Recommendations have been made for an integrated approach that may also include sex therapy, couple therapy and physical therapy (64,67). However, these interventions have not been empirically evaluated for the treatment of postmenopausal dyspareunia.

Nonhormonal vaginal moisturizer
Replens, a polycarbophil-based vaginal moisturizing gel, is endorsed by the Canadian Obstetrics and Gynecology Society as the sole nonhormonal treatment for pain during intercourse  in postmenopausal women (70). This recommendation is based on a comparative, randomized, open-label study which found Replens was as effective as a local estrogen cream in alleviating pain during intercourse in postmenopausal women with vaginal dryness (71). This investigation only assessed dypareunia in participants who engaged in sexual intercourse (ie, n=11 in each condition, pretreatment). At the end of treatment, fewer women were engaging in sexual intercourse than at the outset (ie, Replens n=6, dienestrol n=9), which indicates that some participants experiencing dyspareunia stopped having intercourse and were not considered in the final analysis. At the time of writing, there are no other investigations into the efficacy of nonhormonal vaginal moisturizers for the treatment of postmenopausal dyspareunia.

HRT
Estrogen replacement therapy is thought to offer relief from dyspareunia in postmenopausal women by reversing vaginal atrophy, increasing vaginal blood flow and promoting lubrication during intercourse (72). Both systemic and local administrations are prescribed for vaginal atrophy symptoms, but direct application onto the affected genital area is favoured because it is associated with fewer adverse events and directly targets affected tissue. Current guidelines offered by The Joint Committee -Clinical Practice Gynaecology and Urogynaecology (70) of the Canadian Obstetrics and Gynecology Society state that postmenopausal women experiencing local symptoms of vaginal atrophy, such as dyspareunia, can be prescribed either a conjugated estrogen cream, a sustained-release intravaginal E 2 ring or a low-dose E 2 tablet. Recently, a position statement by the North American Menopause Society and a Cochrane literature review both reiterated these recommendations (73,74). As such, only research into the efficacy of these specifically recommended forms of estrogen replacement is discussed in the present review. Use of other preparations of topical and systemic hormonal therapies (eg, oral formulations of estrogen supplementation with or without progesterone, tibolone, transdermal E 2 ) to alleviate dyspareunia in samples of postmenopausal women, generally suffering from vaginal atrophy, have also been investigated (eg, [75][76][77][78][79][80]. The majority of published clinical studies examining these alternate forms of hormonal supplementation for the treatment of postmenopausal dyspareunia reported positive results regarding mean improvement. However, these preparations cause increased systemic hormonal absorption and are not recommended if urogenital symptoms, such as dyspareunia, are the primary complaint. All reviewed treatment studies demonstrated beneficial effects of recommended estrogen replacement therapies on dyspareunia in postmenopausal women with vaginal atrophy (Table 5). While most were open label, or single-blinded, randomized, parallel-group investigations, three methodologically superior trials used a double-blind, randomized, placebocontrolled design (31,32,81).
In examining randomized parallel-group trials, the continual release vaginal E 2 ring produced response rates (ie, percentage of sufferers reporting symptom improvement) that fell between 75% and 93% and cure rates (ie, percentage of women reporting complete alleviation of symptoms) ranged from 61% to 75% (82)(83)(84)(85).
Randomized parallel-group trials have also shown that conjugated estrogen cream is superior to oral estrogens for the treatment of dyspareunia; it promoted better genital vascularization and produced improvement in 70% of hysterectomized women (86). Furthermore, it was found to be as effective as the vaginal E 2 ring in alleviating dyspareunia in naturally postmenopausal women with vaginal atrophy. Approximately 77% of sufferers reported improvement and 57% reported being cured (83).
A key methodological issue within the reviewed treatment literature is that many clinical trials are not double-blinded, randomized, placebo-controlled trials; they compare outcome to baseline symptom levels. Without a placebo control, the extent of improvement due to placebo effect is unknown (87,88) and can be considerable. For example, three reviewed double-blinded, randomized, placebo-controlled investigations examined recommended forms of localized estrogen therapy (31,32,81) and their positive findings on the beneficial effects of estrogen replacement for postmenopausal dyspareunia are in line with the aforementioned single-blind and parallel group studies. Two of the three studies found that low-dose E 2 tablets improved dyspareunia in 73.1% to 81.2% of sufferers (32,81). The third study demonstrated that treatment with the continual release intravaginal E 2 ring alleviated dyspareunia in 90% of sufferers (31). However, these investigations found that between 27% and 47% of sufferers reported relief of dyspareunia in the placebo control groups (31,32,81).
Although the double-blinded, randomized, placebocontrolled trials were methodologically more rigorous compared with other reviewed investigations, several significant methodological problems should be highlighted. For example, although Simunic et al (32) produced the most comprehensive of these types of studies, the study still suffers from significant methodological problems. Dyspareunia was measured in a categorical fashion (ie, presence versus absence) rather than on a continual scale (ie, none, mild, moderate, severe), which may result in a distortion of reported treatment effects. Furthermore, 45 women in the E 2 treatment group dropped out of the study and an intent-to-treat analysis was not performed; this may have inflated the treatment response rate. Level of sexual activity is often used as a marker of functional improvement in studies examining symptoms that affect sexual functioning, but this was not measured. Also, the statistical analyses used -paired sample t tests -were quite simplistic. These analyses probably inflated type I error and did not take into account the covariance of other vaginal atrophy symptoms or correct for multiple comparisons. Moreover, because this trial only included postmenopausal women with vaginal atrophy, little is known about the efficacy of the studied treatment for dyspareunia that occurs independent of this condition. Given these issues, even though this study is the best clinical trial to date and provides important evidence supporting the use of estrogen, it is not definitive.
It is important to note that all but one of the reviewed treatment studies (86) reporting on the effects of local HRT in postmenopausal dyspareunia exclusively recruited women with vaginal atrophy. However, because evidence for the associations among pain during intercourse, vaginal atrophy and hormone levels is weak, it is possible that these studies do not capture the full scope of dyspareunia occurring after menopause. Although it is estimated that between 20% and 60% of postmenopausal women do not have vaginal atrophy (27,89), there is a paucity of treatment information about postmenopausal dyspareunia that does not occur concurrently with vaginal atrophy. Also, many treatment outcome studies only reported group means, and while improvement was generally reported to be significant, whether this was due to outliers or whether most women  (90). Notwithstanding, between 10% and 27% of dyspareunia sufferers in the most methodologically sound investigations did not improve (31,32,81).

CONCLUSIONS AND FUTURE DIRECTIONS
The present review found evidence that dyspareunia is a common problem in postmenopausal women, frequently related to declining estrogen levels. This is supported by longitudinal and cross-sectional data (59,60). Furthermore, treatment outcome studies of recommended forms of local estrogen administration have demonstrated that they are effective at relieving dyspareunia in most postmenopausal women who have vaginal atrophy (70,73,74). A substantial amount of evidence, however, has been found to suggest that estrogen decline is unlikely to be the sole cause of dyspareunia in postmenopausal women (eg, [61][62][63]. Moreover, a considerable proportion of postmenopausal dyspareunia sufferers with concurrent vaginal atrophy did not benefit from estrogen supplementation (32,81,82,85), and placebo effects seem to be quite strong (31,32,81). Presently, there is a lack of research about factors other than estrogen that may be associated with the development of dyspareunia in postmenopausal women, which may be a result of the current reductionist conceptualization of postmenopausal dyspareunia. Past work from our multidisciplinary research group has initiated a change in perspective towards conceptualizing dyspareunia as a multidimensional pain syndrome (44,91,92). In envisaging dyspareunia within a multiaxial pain syndrome framework, researchers have been guided toward using pain methodology to investigate this condition and its treatment in premenopausal women. This change in approach has had important scientific and clinical implications. For example, we have demonstrated that pain characteristics such as location, temporal pattern, intensity and sensory quality are useful diagnostic criteria to discriminate subtypes of premenopausal dyspareunic pain (44). Furthermore, similar to other chronic pain syndromes (eg, headaches, lower back pain), it has been found that multiple mediating factors contribute to, exacerbate and maintain dyspareunia in premenopausal women; these factors need to be addressed in treatment (eg, 91,93). Indeed, dyspareunia is now classified as a form of chronic pain by the International Association for the Study of Pain (45).
While this paradigm shift has impacted research and treatment of dyspareunic pain syndromes in premenopausal women, it has yet to be extended to analogous pain affecting postmenopausal women. A reconceptualization of postmenopausal dyspareunia as a pain disorder is overdue, which is reflected by the lack of literature concerning dyspareunia. This change in perspective would suggest that the chief symptom, pain, is a function of multiple biological, psychological and interpersonal factors. Similar to other forms of chronic pain, a single etiological mechanism (such as estrogen levels) is unlikely to account for the development and maintenance of all forms of dyspareunic pain in postmenopausal women.