An increasing number of individuals with HIV infection treated with highly active antiretroviral treatment (HAART) have complained about sexual dysfunction (SD) [1–6]. So far, however, in clinical trials SD has not been reported as a potential side-effect associated with HAART, but this is likely to be due to reporting bias, because patients are not routinely or systematically questioned about SD and may be unwilling to admit to such difficulties unless directly challenged. Many reasons why individuals with HIV may suffer from SD can be suggested, but to date only a few studies have addressed the question of whether SD may be caused by HAART. We investigated the prevalence of SD in an HIV-infected population taking antiretroviral treatment, to identify risk factors for developing SD.
A survey was conducted between December 1998 and December 1999 among HIV-infected individuals, using an anonymous questionnaire. Questionnaires were distributed in eight HIV treatment centres and four non-governmental organizations in 10 European countries (see Acknowledgements). SD was evaluated by asking if there was a decrease in sexual interest and sexual potency since commencing antiretroviral treatment. Lipodystrophy syndrome was defined as having at least one of following symptoms: increased abdominal girth, increased breast volume, decreased facial fat or decreased fat over the buttock area. Participants were considered to have peripheral neuropathy when they reported tingling of the extremities. Current and previous use of antiretroviral treatments was expressed as months of exposure to each individual drug. Current therapy was defined as the therapy that was being taken at the time of filling in the questionnaire, and a history of antiretroviral therapy was defined as drugs taken in the past but not included in the current regimen.
Parameters considered as potential predictors for SD were sociodemographic characteristics (age, sex, HIV transmission category), immunological and virological characteristics (CD4 lymphocyte count, time since HIV diagnosis, viral load and clinical stage of HIV disease), previous and current antiretroviral therapy, and non-antiretroviral therapies such as anti-depressants, tranquillisers, anti-epileptic, antihypertensive and lipid-lowering drugs.
Univariate analyses were performed to explore potential covariates. The response variables were either having a decrease in sexual interest or having a decrease in sexual potency. As a measure of correlation, univariate logistic regression was used for a continuous independent variable whereas the Pearson chi-square statistical test was used for a binary independent variable (if the expected cell count was less than five, Fisher's exact test was used). Logistic regression was performed to establish a multivariate model (backward selection with entry level 0.05, and removal level 0.05). P values lower than 0.05 were considered statistically significant. The possible association between SD and other side-effects such as lipodystrophy and polyneuritis was investigated in univariate analyses, but not in multivariate analyses because this would create problems of multicollinearity .
A first multivariate model investigated the correlation between the occurrence of SD and exposure to combination antiretroviral therapy. A second model focused on the relationship between the individual protease inhibitors (PI) and SD. Both models were run twice: once with the antiretroviral therapy coded as a binary variable (treatment yes or no) and once as a continuous variable (number of months on therapy).
All analyses were performed using Statistical Product and Service Solutions (SPSS) version 9.0.
A total of 1026 individuals with HIV infection treated with antiretroviral agents completed the questionnaire. Information from non-respondents was not available. After validating the answers, 904 (88%) questionnaires were included for analyses. The sociodemographic, immunological and virological characteristics of the 122 individuals who were excluded did not differ from the included population. The characteristics of the study participants are shown in Table 1.
A total of 330 participants (37%) complained of having a decrease in sexual interest [279 (38%) of the men and 51 (29%) of the women]. A decrease in sexual potency was reported by 228 men (31%). A decrease in sexual interest was reported significantly more often by PI-experienced patients (308/766, 40%; both men and women), compared with PI-naive patients [22/138, 16%; odds ratio (OR) 3.55; 95% confidence interval (CI) 2.15–5.89]. In addition, a significantly greater number of PI-experienced men reported a decrease in sexual potency (216/628, 34%) compared with PI-naive men (12/99, 12%; OR 2.56; 95% CI 1.33–5.03). Among PI-experienced individuals a decrease in sexual interest was reported more often in individuals with than without signs of lipodystrophy [239 (49%) versus 69 (25%); OR 2.96; CI 2.11–4.15] and in individuals with than without signs of polyneuritis [164 (53%) versus 144 (31%); OR 2.48; CI 1.82–3.39].
In the first multivariate model (Table 2), the following factors were associated with a decrease in sexual interest: a current PI-containing therapy (OR 1.99, 95% CI 1.32–3.00), a history of a PI-containing therapy (OR 1.47, 95% CI 1.05–2.06), symptomatic versus asymptomatic HIV infection (OR 2.54, 95% CI 1.76–3.69), homosexual contact as the transmission mode versus heterosexual contact (OR 0.56, 95% CI 0.36–0.88), and age (OR 1.02, 95% CI 1.00–1.04). Factors associated with a decrease in potency were symptomatic versus asymptomatic HIV infection (OR 2.46, 95% CI 1.60–3.77), age (OR 1.04, 95% CI 1.02–1.06) and tranquillisers (OR 1.88, 95% CI 1.15–3.07). Antidepressants were found to be negatively correlated with a decrease in potency (OR 0.35, 95% CI 0.17–0.71). The current use of a PI-containing therapy was only found to be a predictor of a decrease in sexual potency when a current PI was included in the analysis as a binary variable (OR 2.01; CI 1.22–3.33), but not when therapy was coded as a continuous variable. The duration of HIV infection, CD4 lymphocyte cell count and viral load levels were not found to be associated with SD.
In the second multivariate model (not shown in a table) after correction for therapy history, being on an indinavir (OR 2.56, 95% CI 1.51–4.36) or ritonavir (OR 2.02, 95% CI 1.35–3.01)-containing therapy was found to be associated with a decrease in sexual interest. Indinavir and ritonavir-containing therapies were also found to be associated with a decrease in sexual potency (OR 2.16, 95% CI 1.38–3.40 and OR 2, 95% CI 1.09–3.68, respectively). With therapies re-coded into continuous variables, the same associated factors were identified and no new factors emerged.
The aetiology of SD is often multifactorial, and may be caused by endocrinological, psychogenic, neurogenic, arteriogenic or iatrogenic abnormalities . Numerous medications are also known to cause SD . This study suggests that a decrease in sexual interest and erectile dysfunction are also found in individuals who are on HAART, especially if their regimens contain PI. Among the different PI, indinavir and ritonavir were found to be associated with this impairment of sexual function. Nelfinavir was not found to be associated with SD, but in our study population, the length of time on nelfinavir therapy was significantly shorter than the time on indinavir and ritonavir therapy.
In a recent study performed in the United States, 33% of 217 men who were started on PI treatment presented with SD after a median follow-up of 1.6 years . After adjusting for age and depression, men were 3.4 times as likely to experience impotence after the PI was dispensed than in the period before dispensing.
The finding that most patients who develop SD during PI treatment report a regression or disappearance of this SD after switching to a non-PI treatment regimen also confirms the potential aetiological role of PI for the development of SD .
There are several reasons why few reports exist about SD in the setting of HIV disease or as a complication of HAART regimens. First, patients are often reluctant to talk about sexually related issues with health professionals . In particular, individuals with HIV may be reluctant to ask their doctor how to improve their sex life when they are repeatedly exposed to messages that promote safe sex and reducing the numbers of sexual partners. Physicians, on the other hand, often fail to take sexual histories from their patients and therefore miss dysfunction until it is serious. Finally, quality-of-life questionnaires distributed during clinical trials do not usually contain questions about sexual well being.
The reasons why patients treated with PI develop SD remain unclear. Before the widespread availability of HAART, mainly patients with an advanced stage of HIV infection complained about SD [12–14]. Androgen deficiency was considered to be the main cause of SD in advanced HIV infection . Patients on HAART, on the other hand, report SD despite the fact that their general physical and psychological well being has improved . Neither the use of PI nor lipodystrophy has been associated with significant differences in testosterone, sex hormone binding globulin, prolactin and cortisol levels [16,17]. Erectile dysfunction has been reported with increased frequency by men with hyperlipidaemia, but also during treatment with certain hypolipidaemic drugs (fibrate derivates and statins) . In our study, individuals on a PI-containing therapy who reported having signs of lipodystrophy were more likely to report SD than those who did not have lipodystrophy. It is possible that the SD reported by certain patients with lipodystrophy was related to the fact that they found themselves less sexually attractive, but it is also possible that there is a pathophysiological link between SD and the lipodystrophy syndrome.
In this study, no information was recorded on the psychological well being of the participants. We are therefore unable to determine whether depression might have been a confounding factor. Moreover, the presence of SD could only be compared between individuals taking a PI-containing antiretroviral therapy regimen and those taking a non-PI-containing regimen.
Further research is needed to improve the description of the development of SD in individuals on HAART, to identify pathophysiological mechanisms and to study the management/treatment of this disorder.
The authors would like to thank everybody who filled in the questionnaire. The investigators of the participating centres were: F. Antunes, L. Caldeira (Hospital de Santa Maria, Lisbon, Portugal); A. Castro (Servicio de Medicina Interna, La Coruna, Spain); J. Cobena Guardia (Associacio Anti-SIDA de Catalunya, Spain); F.D. Goebel (Klinikum Innenstadt, München, Germany); J. Kosmidis (General Hospital of Athens, Greece); J. Soletti (AIDES, France), T. Pernetun (Venhälsan, Söderhospital, Stockholm, Sweden); N. Vetter (Pulmologisches Zentrum, Wien, Austria); E.G. Wilkins (North Manchester General Hospital, Manchester, UK).
1. Martinez E, Collazos J, Mayo J. Sexual dysfunction with protease inhibitors. Lancet 1999, 353: 810 –811.
2. Colson A, Sax P. Sexual dysfunction and HIV infection. AIDS Clin Care 2000, 12: 40 –46.
3. Patroni A, Castelli F, Compostella S, El-Hamad I, Airoldi M, Carosi G. Sexual dysfunctions in HAART treated HIV+ patients. XIIIth International AIDS Conference
. Durban, July 2000 [Abstract WeBeB4243].
4. Colson A, Sax P, Keller M, Platt R, Choo P. Sexual dysfunction in protease inhibitor recipients. 7th Conference on Retroviruses and Opportunistic Infections
. San Francisco. February 2000 [Abstract 63].
5. Sollima S, Osio M, Alciati A. Protease inhibitors and male impotence. XIIIth International AIDS Conference
. Durban, July 2000 [Abstract WePeB4216].
6. Colebunders R, Smets E, Verdonck K, Dreezen C. Sexual dysfunction with protease inhibitors. Lancet 1999, 353: 1802. 1802.
7. Neter J, Kutner MH, Nachtsheim CJ, Wasserman W. Applied linear statistical models
. Richard D (editor). Irwin, Illinois: Burr Ridge; 1996. pp. 385 –392 and 410 –416.
8. Wagner G, Sanchez De Tejada I. Update on male erectile dysfunction. BMJ 1998, 316: 678 –682.
9. Finger WW, Lung M, Stagle MA. Medications that may contribute to sexual disorders. J Fam Pract 1997, 44: 33 –43.
10. Colebunders R, Schrooten W, Desmet P, De Roo A, Dreezen C. Adverse effects of antiretroviral therapy. Lancet 2001, 357: 392 –393.
11. Tudiver F, Talbot Y. Why don't men seek help? Family physician's perspectives on help-seeking behaviour in men. J Fam Pract 1999, 48: 47 –52.
12. Tindal B, Forde S, Goldstein D, Ross MW, Cooper DA. Sexual dysfunction in advanced HIV disease. AIDS Care 1994, 6: 105 –107.
13. Dobs AS, Dempsey MA, Ladenson PW, Polk BF. Endocrine disorders in men infected with human immunodeficiency virus. Am J Med 1988, 84: 611 –615.
14. Chalmers A, Catalan J, Nelson M, Day A. Sexual dysfunction in HIV seropositive gay men. VIIIth International Conference on AIDS/3rd STD Congress
. Amsterdam, June 1992 [Abstract PuB 7099, 8:65].
15. Wagner G, Rabkin J, Rabkin R. Illness stage, concurrent medications, and other correlates of low testosterone in men with HIV illness. J Acquir Immune Defic Syndr Hum Retrovirol 1995, 8: 204 –207.
16. Lo J, Mulligan K, Tai V, Algren H, Schambelan M. `Buffalo hump’ in men with HIV-1 infection. Lancet 1998, 351: 867 –870.
17. Bruckert E, Giral P, Heshmati HM, Turpin G. Men treated with hypolipidaemic drugs complain more frequently of erectile dysfunction. J Clin Pharm Ther 1996, 21: 89 –94.
18. Junemann KP, Aufenanger J, Konrad T. et al
. The effect of impaired lipid metabolism on the smooth muscle cells of rabbits. Urol Res 1991, 19: 271 –275.