The effect of male erectile dysfunction on the psychosocial, relationship, and sexual characteristics of heterosexual women in the United States.

Using Internet-based survey data, this study compared the demographic, psychosocial, relationship, and sexual characteristics of three groups of U.S. women: (a) women whose partners had erectile dysfunction (ED) and were taking medication to treat ED; (b) women whose partners had ED in the previous 3 months and were not taking medication to treat ED; and (c) a control group of women whose partners did not have ED. Results indicate that women are affected by their partners' ED and that ED treatment benefits women's sexual self-efficacy, communication about sexual issues, and sexual and relationship satisfaction.

Erectile dysfunction (ED) defined as the persistent inability to achieve or maintain an erection sufficient enough for satisfactory sexual performance, is a highly prevalent disorder in men ([22]). In the population-based Massachusetts Male Aging Study, 52% of men aged 40–70 experienced at least some degree of ED ([6]).

Erectile dysfunction can have a significant and negative impact on men's social and psychological well-being and quality of life ([1]; [18]; [26]). From a clinical perspective, ED has been largely conceptualized and treated as a male medical condition, and as such there has been relatively little research on how ED affects men's partners or their relationships ([24]). Some early research on sexual dysfunction suggests that couples who experience sexual dysfunction may also experience decreased emotional and physical intimacy, and lower satisfaction with their relationship and sexual life ([2]; [9]; [17]; [19]).

The introduction of oral therapies for ED has made it possible for ED to be treated safely, effectively, and conveniently. In addition, ED treatment has been associated with improvement in quality of life among men and their partners, and with improvement in the quality of partner relationships ([10]; [16]; [20]; [21]; [23]). For example, in a questionnaire study following treatment with a phosphodiesterase type-5 (PDE5) inhibitor, [20] reported improved quality of life scores in both patients (51.5%) and their partners (43.1%). In a retrospective survey study of men currently using PDE5 therapy, men reported improvements in satisfaction with their sexual relationship and in their marital/relationship interaction, although the partner perspective was not assessed in the study ([23]).

The objective of the current study was to compare the demographic, psychosocial, relationship, and sexual characteristics of 3 groups of women: (a) women whose partners experienced ED in the previous 3 months and were taking a prescription medication to treat ED; (b) women whose partners experienced ED in the previous 3 months and were not taking a prescription medication to treat ED; and (c) a control group of women whose partners had not experienced ED in the past 12 months.

Clinical studies of ED treatment focus primarily on outcomes such as erectile function and intercourse success in patients with ED. Other potentially important outcomes of treatment for sexual dysfunction include psychological and relationship domains such as relationship satisfaction, sexual satisfaction, and improvement in self-confidence ([3]). The effect of the men's ED treatment status on the psychological and relationship domains, as reported by women, however, has not been well-characterized. For example, would women whose partners had ED that was treated be more similar to women whose partners did not have ED, or would they be more similar to women whose partners whose ED was untreated on psychological and relationship domains? Thus, by comparing women whose partners had treated and untreated ED to those without ED, this article also examined whether ED treatment had some potential psychosocial, relationship, or sexual benefit for women.

This Internet-based survey was administered by Harris Interactive, Inc. (Rochester, NY), a worldwide research and consulting firm specializing in Internet-based methodologies to conduct scientific research. Women in the United States were recruited from the Harris Poll Online panel, a multimillion-member panel of cooperative respondents from more than 100 countries, and from the Chronic Illness subpanel, which includes participants with a number of chronic diseases who are willing to participate in research on health and medical issues.

To participate in the study, respondents had to be female and currently in a sexual relationship with a male partner aged 40–70 years old. The overall study sample was recruited using nonprobability, purposive sampling and random sampling. Our study required 400 respondents whose sexual partner had ED, half of whom were using a prescription medication to treat ED and half of whom were not using medication to treat ED. To achieve the targeted ED partner sample size of 400, e-mail invitations to participate in the study were sent to all women in the Chronic Illness subpanel who had previously indicated that they reside in a household in which a family member has erectile dysfunction (n = 10,300).

In addition, we targeted 150 respondents whose partners did not currently have ED as our control group. To achieve the control group sample size of 150, e-mail invitations to participate in the study were sent to a random sample of 6013 women in the Harris Poll Online panel who had not previously indicated that they had a partner with ED. Once the desired sample sizes in each group were achieved, the data collection was considered complete and the survey was closed for participation.

An e-mail containing a very general overview of the research and an invitation to participate in the study was sent to potential female respondents. The e-mail invitation contained a link and password for accessing the survey online. Passwords assured privacy and prevented the respondents from taking the survey more than once. Respondents were not paid for participation; incentives for participation included points for completed surveys that could be redeemed for a variety of gifts and awards, and inclusion in a sweepstakes for cash prizes. Participants were completely anonymous to the study authors. The study design and confidentiality safeguards conformed to U.S. Federal Trade Commission privacy protection standards. The study design and procedures were approved by the Western Institutional Review Board®. The research met the requirements for a waiver of consent under 45 CFR 46.116(d).

Sexual desire and sexual satisfaction were measured using domains from the Female Sexual Function Index (FSFI), which is a brief, multi-dimensional, self-report measure of female sexual function for clinical and nonclinical samples ([25]). The FSFI is a 19-item measure of the following individual domains for the previous 4 weeks: desire; arousal; lubrication; orgasm; satisfaction; and pain. [25] suggested that the FSFI is most appropriately used in women who have had some type of sexual activity in the past month; therefore, we examined scores only on the desire and satisfaction domains because most of these items were not affected by whether a woman had been sexually active or not and because these were the most relevant domains for a nonclinical sample.

The sexual desire domain consists of two items and the scale range is 1.2–6. The satisfaction domain assesses global sexual and relationship satisfaction and is considered the "quality of life" domain of the scale. This domain consists of three items and the scale range is 0.8–6. Items were scored on a 5-point or 6-point Likert scale depending on sexual activity in the previous month. Item and/or domain scores of 0 on the FSFI indicate that no sexual activity was reported in the previous 4 weeks but do not provide further information about sexual function. Domain scores were computed by adding the item scores and multiplying the sum by a domain factor. Higher scores correspond to greater sexual function. These domains had high internal consistency in this study (α =.88 for desire and α =.90 for satisfaction).

Sexual communication was measured using the Dyadic Sexual Communication Scale (DSC), a 13-item self-report measure of respondents' perceptions of the discussion of sexual matters with their partners ([4]). The DSC is unidimensional scale, with items scored on a 6-point Likert scale (1 = disagree strongly; 6 = agree strongly). Items 1–7 were reverse scored and all items were summed up for a total score of 13–78. Higher scores indicate better communication skills. This scale had high internal consistency in this study (α =.92).

Participants completed the sexual self-efficacy and sexual depression subscales of the Multi-dimensional Sexual Self-Concept Questionnaire, a 100-item, self-report instrument that measures various psychological aspects of human sexuality ([27]). Sexual self-efficacy measures the "belief that one has the ability to deal effectively with the sexual aspects of oneself," and sexual depression measures the "experience of feelings of sadness, unhappiness, and depression regarding one's sex life" ([27], p. 521–522). Each subscale consists of five items scored on a 5-point Likert scale (0 = not at all characteristic of me; 4 = very characteristic of me). Subscale scores represent the average of the item scores, with higher scores representing higher sexual self-efficacy or sexual depression. Internal consistency for both scales was high, α =.94 for efficacy and α =.92 for depression.

Relationship satisfaction was measured with a 7-item measure of general relationship satisfaction, the Relationship Assessment Scale (RAS; [14]). The RAS has been used reliably across samples of ethnically diverse and age-diverse couples, as well as with partners seeking marital and family therapy ([15]). Items were scored on a 5-point Likert scale. The scale score was computed as the mean of the item scores, with higher scores indicating greater relationship satisfaction. The internal consistency of this scale was high, α =.94.

We also asked women several questions about their sexual behavior in the previous month, including how frequently they preferred to have sexual intercourse, the number of sexual intercourse attempts in the previous month, and a question about which partner typically initiated sexual activity.

To examine univariate group differences on variables, we conducted chi-squared analyses for categorical variables (with additional chi-square analyses conducted for specific group differences) and analysis of variance (ANOVA) for continuous variables. We used the Scheffé post-hoc procedure to examine group differences when the overall F-test was statistically significant.

We were particularly interested in understanding the impact of partners' erectile function and treatment status on women's overall relationship satisfaction. Therefore, for this dependent variable we conducted an analysis of covariance (ANCOVA) to control for group differences and to consider the potential interaction of partner ED status and sexual activity on women's relationship satisfaction. We included in the ANCOVA independent variables of partner ED/treatment status and whether women were sexually active in the previous month to determine if there were main and/or interaction effects of these variables on relationship satisfaction. Although groups also differed in women's age and relationship status, these were not included as covariates because women's age was highly correlated with partner age (i.e., it was redundant) and marital status was not correlated with the dependent variable ([28]).

Figure 1 shows how the final study sample was achieved. A total of 16,313 female respondents in the Harris Poll Online (n = 6013) and Chronic Illness panels (n = 10,300) received an e-mail invitation to participate in the study. A total of 2,818 women responded to the e-mail invitation; 2066 did not meet the study entry criteria, 192 did not complete the study questionnaire, and 560 qualified to participate and completed the questionnaire. Subsequently, five participants who completed the questionnaire were found not to have responded to the partner health status question used to determine sample designation and were therefore disqualified from further analyses, leaving a total survey sample of 555. Surveys were completed between November 3, 2003 and November 10, 2003.

Graph: Figure 1 Study sample flow chart.

From the 555 valid participants, we established three groups for the study: women whose partners had recent ED and were taking a prescription medication to treat ED; women whose partners had recent ED and were not taking a prescription medication to treat ED; and a control group of women whose partners had not experienced ED in the past 12 months. To establish the two ED groups, we first selected women who indicated that their partners had experienced ED in the past 12 months (n = 404). From these women, we further refined the sample to those women whose partners recently had ED. Those who responded "yes" to the question, "In the past 3 months, has your partner had consistent difficulty getting or maintaining an erection sufficient for sexual activities," were considered to have partners with recent ED (n = 354). The other 50 women were classified as having partners with nonrecent ED and were not included in this article. Of the 354 women whose partners had ED in the past 3 months, 171 indicated that their partner takes a prescription medication to treat his ED, and 183 indicated that their partner does not take a prescription medication to treat his ED. Women who responded that their partner had not experienced ED in the past 12 months were classified as healthy in terms of their partner's erectile function (n = 151).

The demographic characteristics of the study respondents are summarized in Table 1. In the overall sample, mean age was 53.22 years (SD = 8.50, range 25.00–74.00); partner's mean age was 56.32 (SD = 8.03; range 40.00–70.00). Most of the women surveyed identified their race/ethnicity as white (n = 484; 95.80%). The most commonly reported ED treatment for the women whose partners had treated ED was sildenafil citrate (Sildenafil; Pfizer Inc., New York, NY).

Table 1 Demographic Characteristics of Respondents

There were significant group differences on participant age, F (2, 502) = 5.40, p <.01. Post-hoc analyses showed that women whose partners did not have ED were significantly younger than women whose partners had untreated ED (p <.01). Similarly, there were group differences on male partner age, F (2, 502) = 8.29, p <.01. Post-hoc analyses showed that male partners who did not have ED were significantly younger than male partners who had ED. Relationship status also differed significantly by group, χ2 (2, _I_N_i_ = 505) = 12.63, p =.01. Women whose partners did not have ED were less likely to be married than women whose partners had treated ED, χ2 (1, _I_N_i_ = 322) = 12.36, p <.001.

Group differences on psychosocial outcomes are presented in Table 2.

Table 2 Psychosocial Characteristics of Respondents

We found significant group differences on sexual self-efficacy and sexual depression, F (2, 502) = 4.60, p =.01; F (2, 502) = 18.62, p <.001, respectively). Post-hoc analyses showed that women whose partners did not have ED reported significantly higher sexual self-efficacy scores than women whose partners had untreated ED (p <.05). There were no significant differences on sexual self-efficacy between women whose partners did not have ED and those whose partners were treated for ED, or between women in the two ED groups (i.e., women whose partners had treated and untreated ED scored statistically similarly). For sexual depression, post-hoc analyses showed that women whose partners did not have ED reported significantly lower scores than both groups of women whose partners had ED (p for both groups <.001). There were no significant differences on sexual depression scores between women in the ED groups.

We found a significant group difference on sexual communication, F (2, 502) = 9.29, p <.001. Post-hoc analyses showed that women whose partners did not have ED reported significantly higher sexual communication scores than women whose partners had untreated ED (p <.001). There were no significant differences between women whose partners did not have ED and those whose partners were treated for ED, or between women in the two ED groups.

We found group differences on sexual desire and sexual satisfaction reported in the past 4 weeks, F (2, 502) = 7.54, p =.001; F (2, 502) = 27.96, p <.001, respectively. Post-hoc analyses showed that women whose partners did not have ED reported significantly higher sexual desire scores than both groups of women whose partners had ED (p =.001 untreated and p <.05 treated); partner ED treatment status did not significantly influence sexual desire scores (p =.793). Women whose partners did not have ED reported significantly higher sexual satisfaction scores than both groups of women whose partners had ED (p <.001 untreated and p =.001 treated). Between the two ED groups, women whose partners were treated had significantly higher sexual satisfaction scores than women whose partners were untreated (p =.001).

We found a significant group difference on relationship satisfaction, F(2, 502) = 14.38, p <.001. Post-hoc analyses showed that women whose partners did not have ED reported significantly higher relationship satisfaction scores than women whose partners had untreated ED (p <.001). There were no significant differences between women whose partners did not have ED and those whose partners were treated for ED. Between the two ED groups, women whose partners had treated ED had significantly higher relationship satisfaction scores than women whose partners had untreated ED (p <.01).

Group differences on sexual behavior outcomes are presented in Table 3.

Table 3 Sexual Behavior in the Previous Month

There were significant group differences on sexual activity in the past month, χ2 (2, _I_N_i_ = 505) = 14.12, p =.001. As expected, women whose partners had untreated ED were significantly less likely to report sexual activity in the previous month than women whose partners did not have ED and women whose partners had ED that was treated (p =.001 and p <.05, respectively).

There were significant group differences on the preferred frequency of intercourse, F (2, 502) = 13.64, p <.001, and on the actual number of sexual attempts women had with their partners in the previous month, F (2, 501) = 22.97, p <.001. Post-hoc analyses showed women whose partners did not have ED reported a significantly higher preferred intercourse frequency per month than both groups of women whose partners had ED (p <.001). Partner's ED treatment status did not significantly influence desired intercourse frequency scores among women whose partners had ED. This same pattern across groups was observed among actual sexual intercourse attempts in the previous month.

Groups differed significantly in terms of who typically initiated sexual encounters in the previous month, χ 2 (4, _I_N_i_ = 446) = 22.53, p <.001. Women whose partners had untreated ED were significantly more likely to report that they initiated sexual activity compared to women whose partners had treated ED (p <.01) and women whose partners did not have ED (p <.001). These groups were more likely to report that sex was initiated mutually between partners.

After adjustment for partner age there was a significant main effect for whether or not women were sexually active in the previous month, F (1, 498) = 14.19, p <.001, and a significant interaction between partner ED and treatment status and sexual activity in the previous month, F (2, 498) = 7.02, p =.001. The main effect for partner ED and treatment status was not significant (p >.05) when adjusted for covariates. Adjusted means (Table 4) suggest that among women whose partners did not have ED and women whose partners had treated ED, relationship satisfaction was higher for those women who reported being sexually active in the previous month. Among women whose partners had untreated ED, however, sexual activity did not affect relationship satisfaction scores. Further, relationship satisfaction scores were lowest for women whose partners had treated ED but did not report being sexually active.

Table 4 Mean Scores on Relationship Satisfaction by Partner ED and Treatment Status and Sexual Activity, Adjusted for Partner Age

Few studies have focused on the impact of men's ED or ED treatment on their female partners or couple relationships. In this study we found several differences among the three groups of women in univariate analyses. In terms of demographic characteristics, women whose partners had ED and their male partners tended to be older than their counterparts without ED, which is not surprising as age is closely associated with the development of ED ([6]). Women whose partners did not have ED were also less likely than women in the ED groups to be married, although this was only statistically significantly different from women whose partners had treated ED.

We noted several patterns of differences among the groups on psychosocial dimensions related to sex and relationships. The control group (No partner ED) reported significantly better scores than both of the ED groups on sexual depression, sexual desire, and overall sexual satisfaction. On the measure of sexual satisfaction, however, women whose partners had treated ED had significantly higher scores than the women whose partners had untreated ED. This suggests that men's treatment may have a positive impact on women's sexual satisfaction, but not to the level of women whose partners do not have ED. On the measures of sexual self-efficacy, sexual communication, and relationship satisfaction, the control group differed significantly from the women whose partners had untreated ED but did not differ significantly on these domains from women whose partner's ED was treated. Between the two ED groups, women whose partners had treated ED had significantly higher relationship satisfaction scores than women whose partners had untreated ED, but the ED groups did not differ significantly on measures of sexual self-efficacy and sexual communication, although women whose partners had treated ED reported higher scores on these measures.

A high percentage (> 80%) of all three groups reported being sexually active in the past month, which is not surprising given that one of the study inclusion criteria was that women had to have a current sexual relationship with a male partner. However, a significantly lower percentage of women whose partners had untreated ED reported any sexual activity in the previous month compared to women whose partners had treated ED and women whose partners had no ED. These differences in sexual activity would likely be more pronounced had we not excluded partners who did not have a current sexual relationship. Although most women in the study were sexually active, for the dependent variable of relationship satisfaction we found a significant interaction effect between partner's ED and treatment status and whether women were sexually active in the previous month. Partner's ED and treatment status alone was not significantly related to relationship satisfaction when sexual activity and partner age were covariates in an ANCOVA model.

Taken together, these findings suggest that women are affected by their partners' ED and there may be appreciable ED treatment benefits for women in areas of sexual self-efficacy, communication about sexual issues, and sexual and relationship satisfaction. Several recent studies support our findings. A multi-national survey study of female partners of men with ED ([7]) found that partners of men who were treated for their ED using a PDE5 inhibitor reported better sexual function than women whose partners did not use PDE5 therapy. A Turkish study by [5] compared the sexual function of women whose partners had ED with that of women whose partners did not have ED and found that most sexual function domain scores (i.e., sexual arousal, lubrication, orgasm, satisfaction, pain, and total score) as measured by the FSFI were significantly lower among women whose partners had ED. Further, in this study treatment of men's ED improved women's sexual function in several domains. Although [5] focused specifically on domains of women's sexual function, it provides further evidence that men's ED treatment may provide benefits for women partners in both sexual and psychosocial domains.

Another clinical study ([13]) examined the effect of improved male erectile function on female sexual function and satisfaction outcomes and found that women whose partners received treatment with sildenafil showed greater improvement in sexual satisfaction, enjoyment from sexual intercourse, and several sexual function domains (e.g., arousal, orgasm) compared with women whose partners received placebo.

Finally, a recent clinical study ([7]; [11]) examined sexual function and quality of life outcomes in partners of men with ED who were treated with the PDE5 inhibitor vardenafil HCL (Bayer Pharmaceuticals Corporation, West Haven, CT) or placebo. Partners of men in the treatment group showed significantly greater improvement in sexual quality of life scores and most sexual function domains compared with women whose partners received placebo.

This study used an Internet-based sample of U.S. women who were currently in a sexual relationship, largely caucasian, and married, which may affect the generalizability of the findings to the larger population of women whose partners have ED. However, research suggests that Internet samples are diverse in terms of gender, socio-economic status, geographic region, and age when compared to traditional samples in psychological research and that findings are consistent with other data collection methods ([12]).

In addition, the women in our study represent the first responders to the invitation to participate in the survey and may not be representative of all women whose partners do and do not have ED in our selected age group. In our study we also were not able to compare participants with nonparticipants, which limits our ability to note any systematic differences between those groups.

Finally, this was a cross-sectional study; subsequently, it is difficult to determine causal relationships among erectile function and treatment status, sexual behaviors, and psychosocial variables. We have, however, observed differences in psychosocial and sexual characteristics between women whose partners do not have ED, whose partners have untreated ED, and whose partners have treated ED. The literature on ED and its psychosocial effects from the man's perspective is in its infancy, although the weight of the evidence suggests that ED may result in decreased quality of life and impaired relationships ([1]). We therefore surmise that the respondents' relationship satisfaction may be attributed in part to the partner's erectile function status. Prospective studies will provide more definitive data on the causal relationships among sexual function and interpersonal and psychosocial outcomes for both men and women.

Despite its limitations, this study is among the first to describe the psychosocial, relationship, and sexual characteristics of women whose partners have ED as compared to women whose partners do not have ED. Further, comparisons among groups suggest that the treatment of ED may benefit partners well as men with ED.