Structured review of the use of the Arizona sexual experiences scale in clinical settings

Approximately 40% of women and 30% of men describe sexual dysfunction, although recognition in medical settings is suboptimal, due to problems in reporting and eliciting concerns relating to sexual function and satisfaction. Screening questionnaires may help to support this aspect of clinical practice. The Arizona sexual experiences scale (ASEX) includes items that quantify sex drive, arousal, vaginal lubrication or penile erection, ability to reach orgasm, and satisfaction from orgasm.


| BACKGROUND
Sexual dysfunction and sexual dissatisfaction are common: community surveys find that approximately 40% of women and 30% of men report troublesome sexual dysfunction. Low sexual desire (in women) and premature ejaculation (in men) are the most commonly reported sexual dysfunctions (Laumann, Paik, & Rosen, 1999;Moreira et al., 2005;Moreira et al., 2008). Sexual dysfunctions are closely associated with mental health problems, particularly depressive symptoms, but recognition of sexual dysfunction in primary care is low (Read et al., 1997;Nazareth et al., 2003;Cyranowski et al., 2004;Montejo et al., 2010). Sexual dysfunction increases the risk of depression by 130-200% (Atlantis & Sullivan, 2012;Clayton, El Haddad, Iluonakhamhe, et al., 2014) and 67% of depressed men and 75% of depressed women report problems relating to sexual function (Thakurta, Singh, Bhattacharya, et al., 2012). A positive overall outcome for the patients with major depressive disorder is generally associated with a positive impact on sexual function, although most antidepressant drugs are associated with treatment-emergent sexual dysfunction in many patients (Clayton, Kornstein, Prakash, et al., 2007).
Assessment of sexual dysfunction can be challenging in practice, and differing assessment methods lead to varying findings. For example, an investigation found the prevalence of sexual dysfunction in men is 20% if reliant on spontaneous report by the patient, but 60% after direct enquiry (Segraves & Balon, 2014). Physicians may believe that the affected patients would spontaneously report sexual dysfunction (Clayton et al., 2014), but less than 20% patients who suffer from sexual dysfunction spontaneously report problems or seek help (Moreira et al., 2005;Moreira et al., 2008): "embarrassment" appears to be the main reason for not reporting (Nicolosi et al., 2006). The most reliable approach for assessing the prevalence and nature of sexual dysfunction is a comprehensive, integrated interview that considers multiple dimensions of sexual life (individual, relational, medical, erotic, sexual skill, and situational dimensions) (McCarthy, 2004), but this is time-consuming and often not feasible in practice.
Use of screening questionnaires may be valuable in busy clinical settings. Existing scales fall into different categories: general measures of sexual dysfunction, measures of sexual dysfunction in depressed populations, and female-and male-specific scales (Rizvi, Yeung, & Kennedy, 2011). A scale must demonstrate both reliability and validity in order to prove its acceptability for use. Tests of reliability estimate the ability of an instrument to generate consistent results under the same conditions, whereas tests of validity examine whether the instrument is actually measuring the concept in question (sexual function) and not a co-occurring construct (for example, anxiety) (Rizvi et al., 2011). Use of clinician-administered schedules may increase validity, but self-reporting scales have better reliability, possibly due to the sensitivity of the subject. Screening tools for sexual dysfunction with fewer items are more prone to false positive than false negatives, whereas questionnaires with more items have a better ability to examine specific domains of sexual functioning (Rizvi et al., 2011).
One of the more frequently used scales for assessing sexual functioning is the Arizona sexual experiences scale (ASEX). Its five items quantify sex drive, arousal, vaginal lubrication or penile erection, ability to reach orgasm, and satisfaction from orgasm. It is considered reliable and valid, and has been found concise and easy to administer in clinical settings (Baldwin, Manson, & Nowak, 2015;Clayton et al., 2014;Francois, Levin, Kutscher, et al., 2017;Khin, Kronstein, Yang, et al., 2015;Lorenz, Rullo, & Faubion, 2016;McGahuey, Gelenberg, Laukes, et al., 2000). Possible total scores range between 5 and 30, higher scores indicating greater sexual dysfunction. ASEX designates "sexual dysfunction" when the total score is 19 or higher; or when any single item has a score of 5 or 6; or when any three items each have a score of 4 or higher . The ASEX has been used in many cross-sectional prevalence studies and in randomized controlled trials of pharmacological treatment, but there is some uncertainty about its utility in other clinical settings. We therefore aimed to examine the psychometric properties of the ASEX and its utility in assessing sexual dysfunction in clinical settings.

| RESULTS
We identified 240 records: after excluding letters and duplicates, 228 were screened. Five were excluded as they were not in English, the remaining 223 were assessed for eligibility. From this group, 49 pre-clinical studies were excluded, and another 70 were excluded as they were either not obtainable or irrelevant. The final dataset included 104 papers (Figure 1). The specific features in each category are explained in detail below.
In order to establish its psychometric characteristics, the ASEX authors administered it to 107 control subjects (hospital employees, staff, residents, and faculty members) and 58 psychiatric patients.
Analyses of variance (ANOVAs) revealed significant differences in total ASEX scores between patients and controls (males F 18.1, p < .000; females F = 31.71, p < .000) and between females and males (patients F = 5.22, p = .026; controls F = 5.05, p = .031). Further ANO-VAs revealed significant gender differences for patients on the ASEX items for drive and arousal (F = 4.69, p = .035 and F = 5.88, p = .019, respectively), with a trend on the item for ability to reach orgasm (F = 3.72, p = .059). For controls, there were trends for gender differences on the items for drive, arousal, and ability to reach orgasm (F = 3.57, p = .067; F = 3.51, p = .069; and F = 3.83, p = .058, respectively) . In further exploration by the authors, items on the ASEX correlated with factors and related items on the Brief Index of Sexual Functioning (BISF) (Taylor et al., 2005;Reynolds et al., 1988), but not with depression score, on either the Beck Depression Inventory (BDI) (Beck, Steer, Ball, & Ranieri, 1996), or the Hamilton Depression Rating Scale (HDRS) (Hamilton, 1960) (n = 26) and with "sexual dysfunction" (n = 17). A total ASEX score of ≤11 was found to be the best cut-off point (sensitivity = 100%, specificity = 52%) for screening in this group of patients (Soykan et al., 2004).
A study involving administration of the ASEX to patients with either schizophrenia or schizoaffective disorder (n = 247), found a high degree of agreement between a single-item specific screening question for sexual dysfunction and the ASEX. Overall, the sensitivity (85%), specificity (63.7%), and positive (83%) and negative (67.1%) predictive values for the specific single-item screening question were deemed satisfactory by the authors. In contrast a single-item general side effect question performed rather poorly (sensitivity, 11.3%; specificity, 92.5%; positive predictive value, 76%; negative predictive value, 33%) (Byerly et al., 2006).
Another study in patients with schizophrenia or schizoaffective disorder (n = 137) found a sensitivity of 80.8%, and specificity of 88.1% in relation to the Dickson Glazer Scale for the assessment of Sexual Function Inventory (DGSFI) (Dickson & Glazer, 2000) and AUC = 0.93, with an optimum cut off ASEX total score of 14/15 (Nunes et al., 2009

| Epidemiological studies
Many studies have employed the ASEX to ascertain the prevalence of comorbid sexual dysfunction in samples of patients with a range of psychiatric diagnoses (see Table 1) or physical health problems (see Table 2).
In an investigation among treatment-seeking men with opioid dependence (n = 100), the prevalence of ASEX-determined dysfunction was found to be 48%, and there was strong correspondence with reported dysfunction in at least one of the domains of the International Index of Erectile Dysfunction (IIEF) (Rosen et al., 1997) (92%), and at least one of the domains of the Sexual Functioning Questionnaire Short-Form (CSFQ-14) (90%) (Venkatesh et al., 2014).
A cross-sectional study in women with substance-dependence (n = 213) utilized both the ASEX and non-standardized questions about sexual functioning, along with the Drug Abuse Screening Test (Skinner, 1982), the Short Alcohol Dependence Data questionnaire (Davidson & Raistrick, 1986), and the Fagerström Test for Nicotine Dependence (Heatherton, Kozlowski, Frecker, & Fagerström, 1991): there was a similar prevalence of sexual dysfunction using the two methods of enquiry (Diehl et al., 2016).
Patients undergoing long-term methadone treatment (n = 40), scored significantly higher than controls (n = 40) on the ASEX, the Care and Abuse-Questionnaire (CECA-Q) (Bifulco, Bernazzani, Moran, & Jacobs, 2005), and the Symptoms Check List 90 (SCL-90) (Derogatis & Cleary, 1977). ASEX scores were directly and significantly correlated with CECA-Q neglect score and SCL 90 psychiatric symptoms total score. Methadone dosages were not significantly correlated with sexual dysfunction scores, except for "erectile dysfunction," for which an inverse association was seen. Plasma testosterone levels were significantly lower, but prolactin levels were significantly higher in cases than in controls: levels were significantly inversely correlated with ASEX scores, CECA-Q neglect scores, and psychiatric symptoms (SCL 90 score) among methadone patients. Prolactin levels were directly and significantly correlated with sexual dysfunction scores, psychiatric symptoms at SCL 90, and CECA-Q neglect scores. Neither testosterone nor prolactin levels were correlated with methadone dosage (Gerra et al., 2016).
In a case-control study in patients with alcohol dependence (n = 101), 58.4% of patients had ASEX-determined sexual dysfunction: the highest frequency of dysfunction according to individual items was for arousal (57.4%), followed by problems in desire (54.4%), erection (36.6%), satisfaction with orgasm (34.6%), and ability to reach orgasm (12.87%): patients and controls (n = 50), differed significantly in overall dyadic adjustment, in the domains of dyadic satisfaction and affective expression (Pendharkar et al., 2016).
In an early investigation, the prevalence of ASEX-determined dysfunction in patients with major depressive disorder (n = 514) was 73.4% : ASEX scores were significantly associated with score on the Inventory of Depressive Symptomology, self-report version (IDS-SR) (Rush, Gullion, Basco, Jarrett, & Trivedi, 1996), but not correlated with the Quality of Marriage Index (QMI) (Norton, 1983). An investigation in patients with social anxiety disorder (n = 113, of whom 30.1% had comorbid depression) found the proportion of ASEX-determined sexual dysfunction to be 36.3%: in addition, 36.3% of patients (n = 113) with social anxiety and sexual dysfunction, had a history of childhood physical abuse, and 14.2%, a history of childhood sexual abuse (CSA) (Tekin et al., 2016). In a further investigation of ASEX scores in patients with depression at baseline and 2 months follow up (n = 433), there were marked correlations between the Depression and Family Functioning Scale (DFFS) (Williams et al., 2016); the CGI-S (Zaider, Heimberg, Fresco, Schneier, & Liebowitz, 2003), the Montgomery-Åsberg Depression Rating Scale (MADRS) (Montgomery & Asberg, 1979), the Hamilton Anxiety Scale (HAM-A) (Hamilton, 1959), the Sheehan Disability Scale (SDS) (Leon, Olfson, Portera, Farber, & Sheehan, 1997), the ASEX, the Patient Health Questionnaire-9 (PHQ-9) (Kroenke, Spitzer, & Williams, 2001), and the Work Productivity and Activity Impairment Questionnaire (WPAI) (Reilly, Zbrozek, & Dukes, 1993), (Williams et al., 2016). A further investigation which used two models to elucidate associations between symptoms and CSA found that depression severity and anxious arousal mediated the relationship between CSA and adult sexual function, and anxious arousal and sexual functioning mediated the association between CSA and depressive symptoms; when the models were combined, anxious arousal was the most important mediator of CSA on depression, which in turn mediated associations with adult sexual satisfaction and relationship quality (Dunlop et al., 2015).  (Table 3). A study of patients with chronic hepatitis C (n = 46), found a 35% overall prevalence of ASEX-reported sexual dysfunction, the prevalence being higher in women (50%): the most frequent problems in men were in drive (25%), arousal (17%), and erection (17%);

| Gastrointestinal conditions and relevant interventions
in women, the most frequent problems were in drive (55%), arousal (50%), and reaching orgasm (59%). ASEX scores correlated significantly with age and education. After controlling for other variables, gamma glutamyl transpeptidase (GGT) levels predicted ASEX-scores (Soykan et al., 2005). Another study in patients with chronic hepatitis C (n = 66), who received dietary supplements, found that ASEX scores were more significantly improved with administration of Spirulina platensis than with Silymarin (Yakoot & Salem, 2012). An investigation in women with irritable bowel syndrome (n = 616), randomized to comprehensive self-management or "usual care", found that those meeting ASEX criteria for sexual dysfunction were older, had higher lifetime depression and antidepressant use, more primary care visits, fewer mental healthcare visits, and greater sleep disturbance than those individuals without sexual dysfunction, but there was no significant group differences in gastrointestinal or somatic symptoms. When compared with "usual care" treatment, comprehensive self-management improved the sexual quality of life scores, with a weaker effect on ASEX scores (Eugenio et al., 2012).
Finally, a prospective study found a mean total ASEX score of 20

| Renal disease
The ASEX has been included in a series of investigations of patients with renal disease (Dikici et al., 2014;Hekmat et al., 2016;Koca et al., 2012;Kurdoglu et al., 2012;Ozdemir et al., 2007;Soykan et al., 2005). A point prevalence study in patients with end-stage renal disease (n = 98) determined that 69.4% had ASEX-defined sexual dysfunction (Ozdemir et al., 2007). In a longitudinal investigation among end stage renal disease patients, who had undergone dialysis treatment for at least 12 months (n = 43), ASEX-defined sexual dysfunction was present in 47% at baseline and 42% at 6-month follow-up, and total and item-by-item ASEX scores did not change significantly during this period: in female patients, HDRS (Hamilton, 1960) scores were significantly higher in patients with ASEX-reported sexual dysfunction, both at baseline assessment and at follow-up (Soykan et al., 2005). A study of female patients undergoing long-term haemodialysis (n = 140) found significant correlations between total ASEX score, age and duration of haemodialysis, though no correlation between serum haemoglobin, parathyroid hormone, creatinine, iron, calcium, phosphorus, and urea reduction ratio and the ASEX score: there was also a significant difference in total ASEX score between cases and controls (Hekmat et al., 2016). Another investigation in haemodialysis patients (n = 246) found higher ASEX scores among patients with comorbid restless legs syndrome (RLS) than those without RLS, and significant relationships between ASEX scores and demographic variables including educational achievements, occupation, and marital status (Dikici et al., 2014). An investigation in female patients who underwent renal replacement therapy found that ASEX-determined sexual dysfunction rates were significantly higher in a haemodialysis group (n = 39) compared to the peritoneal dialysis group (n = 43) and the kidney transplant group (n = 33), and sexual dysfunction rates were higher in kidney transplant and dialysis patients when compared with controls: multivariate analysis indicated that marital duration and haemodialysis were independent risk factors for sexual dysfunction in the renal replacement population (Koca et al., 2012). A further investigation found that total ASEX scores, ability to reach orgasm, and BDI scores were significantly higher among peritoneal-dialysis (n = 22) and haemodialysis (n = 25) patients than controls (n = 30): peritonealdialysis patients with depressive symptoms were 24 times more likely to experience sexual dysfunction than those without depression, and serum FSH and LH levels were positively correlated with arousal and erection/lubrication scores in the depressed peritoneal-dialysis patients (Kurdoglu et al., 2012).
An early investigation in female patients with vitiligo (n = 50) or chronic urticaria (n = 50) found that ASEX total scores were significantly higher than in controls (n = 50): sexual drive and satisfaction item scores were significantly lower in both patient groups, female patients had more dysfunction in reaching orgasm, and male patients reported less orgasm satisfaction (Sukan & Maner, 2007). A healthycontrol study in patients with neurodermatitis (n = 31) or psoriasis (n = 24) found that neurodermatitis patients reported more sexual (ASEX-defined) and depressive symptoms (assessed with the BDI), than patients with psoriasis or healthy controls (n = 33) (Mercan et al., 2008). An investigation in patients with hidradenitis suppurativa (n = 300) found that female sex and later age of onset were both associated with poor sexual function: poor quality of life was associated with anogenital involvement, early age of onset, and disease severity; whereas sexual health was positively associated with quality of life in female but not male patients (Janse et al., 2017). Finally, ASEX scores were significantly higher in male patients with genital warts (n = 116) than controls (n = 71), there being positive correlations between BDI and BAI scores with ASEX total and item scores (Kucukunal et al., 2013).

| Malignancies
A series of investigations of sexual function have employed the ASEX to assess sexual function among groups of patients with various forms of malignancy (Mathias et al., 2006;Cleary et al., 2011;Yilmaz et al., 2015;Batıo glu-Karaaltın et al., 2017;Surbeck et al., 2015). A descriptive, correlational study in women with reproductive system malignancies (cervical, ovarian, endometrial, and vulvar) (n = 106), found higher ASEX scores 6 weeks post-diagnosis (Cleary et al., 2011). An investigation in patients with laryngeal carcinoma (n = 74) found that ASEX scores in total or partial laryngectomy patients were not significantly different (13.98 ± 6.32 and 13.08 ± 4.96, respectively), though mean BDI scores were significantly higher in total laryngectomy patients (13.20 ± 10.41 vs. 7.76 ± 8.14): BDI scores correlated with Rosenberg Self-Esteem Scale (RSES) (Rosenberg, 1965) scores, and ASEX scores correlated with age (Yilmaz et al., 2015). In a second investigation in patients with laryngeal carcinoma patients, 90.3% of total laryngectomy patients and 63.9% of partial laryngectomy patients had experienced negative effects on sexual function, and ASEX scores were correlated with average scores on the sexuality sub-unit (QL-35 59-60) of the Cancer and Head and Neck module (Sherman et al., 2000), (Batıo glu-Karaaltın et al., 2017). An investigation in patients who underwent surgical resection for diffuse low-grade glioma (n = 32) found that ASEX-determined sexual dysfunction was present in 44% of patients (60% of women, 29% of men), and 53% reported post-operative changes in sexual function (with deterioration in 88%, but improvement in 12%). Right-sided resections were associated with more dysfunction in reaching orgasm than left-sided resections, temporal lobe resection was linked to lower sexual drive and sexual arousal in men than in women, and continued antiepileptic drug treatment in patients who underwent right-sided resection was associated with higher ASEX scores in men than women (Surbeck et al., 2015).
An investigation of the effects of 8 weeks of bupropion (150 mg/day) treatment on sexual function in breast cancer patients who had undergone chemotherapy but were currently receiving radiotherapy (n = 20) found mean ASEX scores declined from 23.45 at baseline to 18.95 at endpoint (Mathias et al., 2006).

| Non-malignant gynaecological or postmenopausal conditions
A number of studies have employed the ASEX to assess sexual dysfunction in patients with a range of conditions, (Kovalevsky et al., 2010;Bachmann et al., 2010;Veras et al., 2011;Portman et al., 2014;Senturk et al., 2015;Pinkerton et al., 2016). An investigation in women with polycystic ovarian syndrome (n = 88) found a mean ASEX score of 14.4 and an overall prevalence of sexual dysfunction of 13.3%: with negative correlations between the ASEX scores and the levels of total testosterone, luteinizing hormone and dehydroepiandrosterone sulfate (Veras et al., 2011). An investigation among postmenopausal women (n = 229) found a mean ASEX score of 19.97, with a positive correlation to the mean total score on the Menopause Rating Scale (Hauser, Huber, Keller, Lauritzen, & Schneider, 1994), (Senturk et al., 2015). In a randomized, double-blind, placebo-controlled study involving post-menopausal patients with symptoms of moderate or severe vulvar and/or vaginal atrophy (n = 652), treatment with bazedoxifene /conjugated oestrogens was associated with a significantly greater improvement from baseline to endpoint in ASEX lubrication item score from baseline, when compared with placebo (but with no significant difference in change in total ASEX score). There were also significant advantages over placebo in vasomotor function, sexual function and total scores on the Menopause-Specific Quality of Life questionnaire (QLS) (Hilditch et al., 1996)

| Neurological conditions
The ASEX has been used to determine the prevalence of sexual dysfunction in patients with Parkinson's disease (Celikel et al., 2008;Jitkritsadakul et al., 2015;Özcan et al., 2015) and multiple sclerosis (Celik et al., 2013). A case-control study among patients with Parkinson's disease (n = 45) found that female patients had reduced sexual drive and were less satisfied with orgasm than controls, whereas male patients reported easier orgasms than controls: regression analysis identified increased age and female sex as predictive of reduced sexual drive and sexual arousal (Celikel et al., 2008).
In an investigation of sexual function which found the point prevalence of ASEX-determined dysfunction to be 81.6% in Parkinson's patients (n = 60) compared to 48.3% of controls, ASEX score was correlated with disease severity and depressive symptoms: logistic regression analysis found factors related to sexual dysfunction included absence of recent sexual intercourse, postural instability, and HAMD item 14 (sexual symptoms) (Jitkritsadakul et al., 2015).

In a separate investigation in patients with
Parkinson's disease (n = 89) which found a mean ASEX total score of 18.54 (SD ±7.27), ASEX total scores were correlated with age, disease stage and HAMA scores: there was no correlation between disease duration and ASEX item scores, but motor symptom scores were correlated with dysfunction in erection or lubrication, HAMD score with orgasm dissatisfaction, and HAMA score with dysfunction in stimulation and orgasm (Özcan et al., 2015). An investigation of sexual function among patients with multiple sclerosis (n = 89) found that women reported ASEX-arousal difficulties significantly more than frequently than men (7.9% versus 1.1%): women also had significantly higher scores on the Multiple Sclerosis Intimacy and Sexuality Questionnaire-19 (MSISQ-19) (Sanders, Foley, & LaRocca, 2000) than men (42.6 ± 12.9 vs. 36.6 ± 13.3) (Celik et al., 2013).

| Cardiovascular function
In a study in male patients with major depressive disorder (n = 46), regression analysis indicated that ASEX scores were predicted by greater

| Other medical conditions
In male patients with a pelvic fracture (n = 40) but without consequent vascular, neural or urogenital problems, sexual dysfunction was infrequent (10%) (Copuroglu et al., 2017). A case-controlled investigation of a mixed group of patients with migraine or tension-type headache (n = 74) found that ASEX items 1-4 were all significantly higher than in migraine patients than controls, and ASEX total and item scores were higher in patients with tension-type headaches than controls: there were no significant relationships between headache features and ASEX score in either group of patients (Bestepe et al., 2011). Finally, a case-controlled investigation in patients with Behçet's disease (n = 50) found that ASEX, HDRS, HARS and Golombok Rust Sexual Satisfaction Scale (GRISS) (Rust & Golombok, 1986) scores, were all significantly higher in patient group than controls (n = 50) (Gül et al., 2013).
In a double-blind, randomized, placebo-controlled treatment study in depressed patients (n = 342), the incidence of ASEXdetermined treatment emergent sexual dysfunction was found to be 46.5% of patients who received duloxetine, compared to 62.8% with paroxetine and 40.5% with placebo. The same study found that ASEX-determined sexual dysfunction rates after 8 weeks (n = 256), were 21.4% for duloxetine, 21.6% for paroxetine and 37.9% for placebo (Detke et al., 2004). Further studies with duloxetine found noninferiority versus paroxetine for ASEX-determined sexual dysfunction (n = 392) (Perahia et al., 2006), and the probability of emergent sexual dysfunction of 49.6% in non-responders and 33.2% in responders during initial treatment (60-120 mg/day for up to 34 weeks, n = 514): treatment responders (n = 288) were randomly assigned to receive either duloxetine or placebo during a further 52-week double-blind maintenance phase, and there was no difference in ASEX score between the placebo and duloxetine groups . In a comparative study in patients with major depressive disorder but without baseline sexual dysfunction (n = 1,647), the prevalence of ASEX-determined dysfunction after 6 months of treatment was similar with duloxetine (23.4%) and SSRI monotherapy (28.7%) (Dueñas et al., 2011).
A double-blind, randomized study in patients with major depressive disorder (n = 323), who received either escitalopram (10-20 mg/ day) or paroxetine (20-40 mg/day), found a high prevalence of ASEX-determined dysfunction at baseline, with a slight increase in ASEX scores above baseline values during acute treatment in both groups, but a subsequent slight decline below baseline values towards the end of maintenance treatment (Baldwin et al., 2006). A randomized, placebo-controlled study in patients with major depressive disorder (n = 410), found no significant differences in ASEX scores between patients allocated to vilazodone or placebo, in men or women (Rickels et al., 2009); these findings being repeated in a further study, that found no significant difference between placebo and vilazodone in ASEX-defined sexual dysfunction (Khan, 2009 (Williams, Abraham, & Symonds, 2010;Williams, Edin, et al., 2010).
A 12-week double-blind, randomized placebo-controlled study of desvenlafaxine (50 mg/day) in patients with major depressive disorder (n = 422) found no significant adverse effect on sexual function (with the exception of orgasmic dysfunction in men without pre-existing sexual dysfunction). Greater orgasmic dysfunction at Week 12 was observed in the sub-group of men without baseline sexual dysfunction treated with desvenlafaxine, relative to placebo. Conversely, women without baseline sexual dysfunction experienced poorer overall sexual functioning and orgasm satisfaction at Week 12 with placebo, compared to desvenlafaxine. Sub-group analyses of treatment responders and non-responders found no difference in the proportion of men or women who developed or had resolution of sexual dysfunction in the desvenlafaxine and placebo groups (Clayton et al., 2013). In a further analysis of the incidence of sexual dysfunction during desvenlafaxine treatment, rates of ASEX-determined dysfunction were comparable with different doses of desvenlafaxine, and comparisons for desvenlafaxine versus placebo of change from baseline in ASEX total score and individual item scores were not significantly different, neither was there a significant treatment-by-gender interaction (Clayton et al., 2015).
A small study in patients with major depressive disorder (n = 33) receiving SSRI or placebo found a prevalence of ASEX-determined dysfunction of 73.7% with SSRIs and 85.7% with treatment-free controls, but no significant differences between groups: dysfunction was associated with female gender, regardless of treatment (Tufan et al., 2013). A survey in patients with depressive and anxiety disorders An 8-week randomized placebo-controlled study of vortioxetine in patients with major depressive disorder (n = 469) found that ASEX total scores during treatment were similar, with no significant differences in depressive symptoms . In a comparison of vortioxetine (2.5 or 5.0 mg), duloxetine (60 mg) and placebo, rates of ASEX-determined sexual dysfunction were 51.0%, 37.5%, 46.9%, and 33.3% in the vortioxetine 2.5 mg, vortioxetine 5 mg, duloxetine, and placebo groups, respectively (Mahableshwarkar et al., 2013).
Antidepressant drugs are often used in patients with diagnoses other than depressive illness. A double-blind placebo-controlled study of fluvoxamine controlled-release in patients with generalized social anxiety disorder (n = 300), found that fluvoxamine did not cause ASEX-determined sexual dysfunction (Westenberg et al., 2004). In patients with generalized anxiety disorder, sexual dysfunction was present in 50% of patients treated with paroxetine when combined with placebo, compared to 38% of patients treated with paroxetine when combined with mirtazapine (Schutters et al., 2011). A doubleblind randomized, controlled study of sublingual alprazolam tablets in acute treatment of patients (n = 190) with panic disorder found that there was no improvement in ASEX scores, despite improvements in scores on the Clinical Global Impressions (CGI-S/CGI-I) (Guy 1976;Busner & Targum, 2007), HAMA, Patient Global Impression (PGI) (Guy, 1976), Psychological General Well-Being Index (PGWBI) (Dupuy, 1984),and Panic Disorder Severity Scale (PDSS) (Shear et al., 1997) (Márquez et al., 2011). In a retrospective cohort study in patients with comorbid binge eating disorder and major depressive disorder, bupropion was superior to sertraline in reducing weight and improving ASEX scores (Calandra et al., 2012). In a double-blind, randomized placebo-controlled study in patients (n = 781) with primary generalized anxiety disorder, rates of treatment-emergent sexual dysfunction in vortioxetine-treated groups were similar to those with placebo (Mahableshwarkar et al., 2013). Some studies have employed the ASEX to investigate the management of sexual dysfunction in patients receiving antidepressants.
Switching to tianeptine (n = 23) resulted in a significant difference between baseline and week 4 or week 8 in ASEX scores, associated with significant improvement in HAM-D scores (Atmaca et al., 2003).
A 6-week randomized controlled trial (n = 101) found that aripiprazole augmentation and antidepressant switching had comparable effect on sexual dysfunction, as assessed by ASEX scores (Han et al., 2015).
Switching to agomelatine in patients with acute depressive episodes (n = 25) led to improved ASEX scores after 3 weeks of treatment (mainly in women rather than men): visual analogue scales for desire, arousal, time, and intensity of orgasm and vaginal lubrication showed improvement in all stages of sexual response in women, with minimal changes in men, and treatment was associated with reduction in depressive symptoms (Sapetti, 2012). The 5-HT 1A agonist and 5-HT 2A antagonist flibanserin was found to be associated with low rates of ASEX-reported treatment-emergent sexual dysfunction in women with major depressive disorder (n = 523): 70% of flibanserin-treated women with baseline sexual dysfunction reported an improvement in sexual function, compared with 30% of placebo-treated women (Kennedy, 2010). Switching to mirtazapine (open-label) for up to 6 weeks (n = 19), led to a return of normal sexual functioning (assessed by ASEX) in 58% of patients, and 11% reported a significant improvement in sexual functioning . Treatment augmentation with Maca 3.0 g/day (a Peruvian plant), (n = 20) had a significant improvement in ASEX and Massachusetts General Hospital Sexual Function Questionnaire (MGH-SFQ) (Dording et al., 2008).
Augmentation with VML-670 (a 5-HT 1A receptor agonist), (n = 88) conferred no significant advantage over placebo (Baldwin et al., 2008). Augmentation with bupropion (n = 41) (DeBattista et al., 2005), (n = 30) (Masand, Ashton, Gupta, & Frank, 2001) or methylphenidate (Pae et al., 2009), also had no significant advantage on ASEX-scores. A placebo controlled trial investigated the use of kavalactones in patients with generalized anxiety disorder (n = 75). Kavalactone administration significantly increased ASEX-sexual drive scores in female participants when compared to placebo, with no negative effects seen in male participants: and there was a highly significant correlation between ASEX reduction (improved sexual function and performance) and anxiety reduction in the whole sample (Sarris et al., 2013).
An investigation involving drug-naïve patients with major depressive disorders (n = 56), found that the 5-HT-2A receptor −1,438 AA genotype was significantly over-represented among the sub-group of patients who experienced sexual dysfunction during SSRI or venlafaxine monotherapy: mean baseline HAMD-17 score, mean baseline ASEX score, and mean end-point ASEX score were all significantly higher than in patients without sexual dysfunction, although mean end-point HAMD-17 scores did not differ significantly between groups (Liang et al., 2012). Another study in patients with paroxetine-induced sexual dysfunctions (n = 55), found a significantly higher rate of ASEX-determined dysfunction among females with a "poor" metabolic status-CYP2D6-phenotype (Zourková et al., 2007).
An early small case series (n = 8) involving a switch to quetiapine in patients with sexual dysfunction associated with previous antipsychotic treatment found a clinically and statistically significant improvement in ASEX total score, significantly decreased total score on the Positive and Negative Syndrome Scale (PANSS) (Kay, Fiszbein, & Opler, 1987), and decreased plasma prolactin levels after transition to quetiapine (Byerly et al., 2004). A subsequent study in patients with schizophrenia (n = 36), found a statistically significant increase in mean ASEX score after 4 weeks of treatment with quetiapine, compared with scores at baseline: the most frequent dysfunction was diminished sexual desire, both in men (31.8%) and women (28.6%), but there was no significant correlation between ASEX scores and plasma prolactin levels (Atmaca et al., 2005). A cross-sectional survey in symptom-remitted male and female patients with schizophrenia receiving antipsychotic medication (n = 827) found that 52.6% had ASEX-determined dysfunction (54.2% reported low sexual desire and 41.7% reported problems in orgasm). In men, erectile dysfunction and ejaculatory problems were seen in 48.1% and 64.2% respectively, and amenorrhea was seen in 24.9% of women. ASEX scores were affected significantly and independently by disease severity in men; ASEX scores were higher in cigarette smokers; low sexual desire was more prevalent among women prescribed first-generation drugs; men undergoing second generation antipsychotic monotherapy had lower ASEX scores than men undergoing combination treatment; and men undergoing combination therapy had more ejaculation problems (Uçok et al., 2007).
A small (n = 22) randomized comparator-controlled investigation of the relationship between prolactin and sexual function in outpatients undergoing treatment with risperidone or quetiapine found that higher serum prolactin levels were related to greater ASEXdetermined sexual dysfunction in men treated with risperidone, but not with quetiapine (Nakonezny et al., 2007). A small (n = 42) randomized double-blind study of either continued risperidone or switch to quetiapine in patients with risperidone-associated sexual dysfunction found no significant treatment effect for either ASEX total score or ASEX items, and no significant treatment x time interaction for either ASEX total scores or ASEX items (Byerly et al., 2008). A large (n = 555) comparative investigation in patients with schizophrenia found a significantly greater improvement from baseline in ASEX score with aripiprazole than with comparator drugs: in addition, although serum prolactin levels were similar in the treatment groups at baseline, mean decreases in serum prolactin were 34.2 mg/dL with aripiprazole, compared with 13.3 mg/dL with comparators (Hanssens et al., 2008). A study in male patients with schizophrenia undergoing antipsychotic drug treatment (n = 89) found that ASEX scores were significantly higher in patients on risperidone, compared to patients on olanzapine: sexual dysfunction and treatment non-adherence were not related to either prolactin or gonadal hormone levels (Konarzewska et al., 2009). An investigation of augmentation with cabergoline in patients (n = 80) receiving a range of antipsychotics found a reduction in prolactin levels with cabergoline treatment in all patients, with mean levels of 73.3 ng/mL (±46.8) to 42.0 ng/mL (±27.8) at month 3 and 27.1 ng/mL (± =20.4) at month 6: mean total ASEX scores also declined, from 19.1 (±5.1) to 17.6 (±5.5) at month 3, and 15.0 (±6.5) at month 6 ( Kalkavoura et al., 2013). Finally, a randomized, double-blind, crossover, placebo-controlled investigation of lodenafil in male patients (n = 50) with schizophrenia and erectile dysfunction found that both lodenafil and placebo were associated with an improvement in ASEX, IIEF scale, PANSS, and QLS scores, with no statistical differences between treatment groups in all sexual domains or in hormone levels (Nunes et al., 2013).

| Primary sexual dysfunction
The ASEX has been used to study primary sexual dysfunction and the potential benefit of pharmacological agents to modulate sexual drive and function (  (Aydogan et al., 2012).
A randomized controlled trial, studying placebo and nocebo responses in heterosexual men (n = 48), found increased levels of sexual function after administration of cabergoline with significant effects for several parameters measured by ASEX, and the Acute Sexual Experience Scale (ASES) (Krüger et al., 2003). Placebo effects were induced only to a small degree. No negative effects on sexual parameters in the nocebo condition were noted. This paradigm could induce only small placebo and nocebo effects (Kruger et al., 2016). Intranasal oxytocin administration to healthy participants (n = 58), did not alter "classical" parameters of sexual function, such as sexual drive, arousal or penile erection and lubrication. However, analysis of variance and a hierarchical linear model found that oxytocin increased the intensity of orgasm andcontentment after sexual intercourse, these effects being more pronounced in men. Men additionally indicated higher levels of sexual satiety after sexual intercourse with oxytocin administration. Women felt more relaxed and subgroups indicated better abilities to share sexual desires or to empathize with their partners (Behnia et al., 2014).
T A B L E 5 ASEX rated primary sexual dysfunction and treatment response A randomized double-blind crossover in women with "hypoactive sexual desire disorder" (n = 10), found significantly improved ASEXitem 2 (arousal) scores on testosterone gel, compared to placebo: similar trends were found in Sexual Function Questionnaire (SFQ-V1) (Quirk et al., 2002) scores (Chudakov et al., 2007).
A cluster-analytic study of use of pornography (n = 875), found that recreational users reported higher ASEX-satisfaction and lower sexual compulsivity, avoidance, and dysfunction, whereas users with a compulsive profile presented lower ASEX satisfaction and dysfunction and higher sexual compulsivity and avoidance. Highly distressed less active users were sexually less satisfied and reported less sexual compulsivity and more sexual dysfunction and avoidance. A larger proportion of women and of dyadic users was found among recreational users, whereas solitary users were more likely to be in the highly distressed less active profile, and men were more likely to be in the compulsive profile (Vaillancourt-Morel et al., 2017).

| Sexual dysfunction with non-psychotropic medications
Few studies have used the ASEX to determine potential dysfunction associated with non-psychotropic medication (  (Chiriacò et al., 2016). Healthy young men without any baseline sexual dysfunction (n = 54), who were taking finasteride for male pattern hair loss, had ASEX-reported sexual side effects associated with finasteride (89%), after 9-16 months (mean 14 months). Neither the length of finasteride use nor the duration of the sexual side effects correlated to changes in scores of sexual dysfunction. Persistent sexual side effects (≥3 months) continued for many months or years despite the discontinuation of finasteride (Irwig, 2012). Another study in women with overactive bladder (n = 30) found that tolterodine immediate release improved mean total ASEX scores compared to baseline: mean item scores for sexual desire, arousal, vaginal lubrication, orgasm, and orgasm satisfaction all improved significantly at follow-up (Hajebrahimi et al., 2008). which suggested that distinction between the two was artificial (Graham, 2010); and the rejection of a linear model for arousal (Graham, Sanders, Milhausen, & McBride, 2004;Janssen, McBride, Yarber, Hill, & Butler, 2008). The five items of the ASEX, however, distinguish between sexual drive and sexual arousal.
The recent criteria, by DSM-5, for the diagnosis of sexual dysfunction requires the presence of clinically significant distress along with sexual difficulties. It is to be mindful that findings from some of the studies included in this review used the term "sexual dysfunction" broadly and not in accordance or predated, the up-to-date definition of the clinical phenomena (according to . Therefore, some of the studies included give evidence for the presence of sexual difficulties.
The ASEX scale appears to have excellent internal consistency and scale reliability and strong test-retest reliability: furthermore, ASEX scores appear to correlate well with factors and related items on other validated questionnaires for assessing sexual dysfunction.
ASEX has very high sensitivity and specificity, and very high positive and negative predictive values. ASEX is available in 43 languages.
Studies which have examined the psychometric properties of the illnesses. The psychometric properties of ASEX in cohorts with anxiety disorders were largely uncertain, and little is known about possible relationships between peripheral biological markers and ASEX scores.
Most studies have found higher rates (48-58.4%) of ASEXreported sexual dysfunction in patients with substance misuse when compared with controls. Patients with MDD were found to have elevated rates of sexual dysfunction (73.4% in mixed samples), and ASEX-reported dysfunction was found to correlate positively with the presence and severity of affective symptoms in patients with MDD. A small group of studies have found no significant correlation between ASEX scores and plasma prolactin levels with quetiapine or aripiprazole in patients with schizophrenia.
Many studies have employed the ASEX to investigate treatmentemergent sexual dysfunction, principally with antidepressant and antipsychotic drugs. Between 32 and 73% of patients receiving antidepressant treatment were found to have sexual dysfunction. Studies which compared groups based on response to treatment found sexual dysfunction in 49.6% in treatment non-responders and 33.2% in treatment responders, with an improvement percentage (IP) −1.6% on ASEX total score in treatment responders. Acute treatment with paroxetine was found to have higher rates of sexual dysfunction (62.8%).
Head-to-head comparative studies have generally found no inferiority between different antidepressants in the sexual dysfunction domain.
Studies have found beneficial effects for aripiprazole augmentation, for switching to agomelatine, and for switching to mirtazapine. 52.6% of patients receiving antipsychotics were found to have sexual dysfunction, and a significantly higher ASEX total score was seen in patients receiving risperidone, compared to patients receiving olanzapine, quetiapine or aripiprazole. Augmentation with cabergoline was associated with improved ASEX scores in patients undergoing antipsychotic treatment.
Studies have found sexual dysfunction in 47-69.4% in patients with renal disease, dysfunction being more pronounced in patients receiving peritoneal dialysis, when compared to those receiving haemodialysis. Studies in patients with malignant conditions have found a significant increase in ASEX scores when compared with controls, post-diagnosis and after cancer treatment. Patients with a number of skin conditions (hidradenitis, genital warts, psoriasis, vitiligo, and neurodermatitis) were found to have a higher incidence of sexual dysfunction when compared with controls, and patients with some movement disorders (parkinsonism, multiple sclerosis) had sexual dysfunction more frequently than controls.
Little is known about the utility of ASEX in anxiety disorders and psychogenic sexual dysfunction. Future research should also look into the utility of ASEX for patients undergoing psychological interventions and couple therapies. Future studies should also include investigation of response shift.