Sexual Problems Among Women and Men Aged 40–80 y: Prevalence and Correlates Identified in the Global Study of Sexual Attitudes and Behaviors

EO Laumann; A Nicolosi; DB Glasser; A Paik; C Gingell; E Moreira; T Wang; for the GSSAB Investigators' Group


Int J Impot Res. 2005;17(1):39-57. 

In This Article


The GSSAB is the first large, multi-country survey to systematically study attitudes, beliefs, and health in sexual relationships in middle-aged and older adults. The survey involved 13 882 women and 13 618 men, aged 40–80 years, in 29 countries, representing many world regions. In Europe (Austria, Belgium, France, Germany, Italy, Spain, Sweden, and the United Kingdom), North America (Canada and USA), Australia, New Zealand, Israel, and Brazil, samples were based on random-digit-dialing and respondents were selected randomly within households by asking for the person between 40 and 80 y of age with the most recent birthday. Sampling in Middle Eastern countries (Algeria, Egypt, Morocco, and Turkey) employed a door-to-door protocol, where households were selected using random starting points, and the study staff contacted every third house in several major cities. In Asian countries (China, Hong Kong (although Hong Kong is part of China, it is listed separately because of its distinct socioeconomic and cultural characteristics), Korea, Indonesia, Malaysia, Philippines, Singapore, Taiwan, and Thailand), an intercept protocol was employed in major cities. Both the door-to-door and intercept protocols represent accepted survey methods for each country, but are likely to be more reflective of their urban populations. The standard sample size of 1500 (equal numbers of men and women) was used in Germany, Sweden, UK, France, Italy, Spain, Australia, Turkey, and Japan. A sample size of 500 was used in Austria, Belgium, New Zealand, Algeria, China, Hong Kong, Taiwan, Indonesia, Malaysia, Philippines, Singapore, and Thailand. In the remaining countries, the sample sizes were: Israel 505, Canada 1007, South Africa 999, USA 1491, Brazil 1199, Mexico 506, Egypt 584, Morocco 509, and Korea 1200.

In Europe, North America, Australia, New Zealand, Israel, and Brazil, telephone interviews were conducted via Computer-Assisted Telephone Interview (CATI). Due to the sensitive nature of the topic, refusals were not called back. The door-to-door and intercept protocols employed in Algeria, Egypt, Morocco, Turkey, China, Korea, Taiwan, Indonesia, Malaysia, the Philippines, Singapore, South Africa, and Thailand used self-completed questionnaires. There were two exceptions to the above-mentioned data-collection strategy. In Japan, a mailed, self-completed questionnaire was used, and in Mexico, mixed-mode method of in-person and telephone interviews was employed. The mean overall response rate was 19%, with the mean rate for the telephone interviews at 15% and 30% for the self completed questionnaires and 33% for the mailed, self-completed questionnaires used in Japan. Response rate ranged from 8–55% in the various countries.

Verbal consent was obtained from all study participants. They were also informed about the following issues: (1) all information obtained would be used in aggregate, (2) responses were voluntary, (3) the confidentiality and the privacy of their responses were protected because no personal identifiers were coded into the interview instruments, (4) no list of respondents was retained, and (5) 'refusers' were not called back in an effort to convert them to participating respondents.

The questionnaire asked for information about demographics; health; relationships and general satisfaction with life as a whole; as well as individual behavior, practices, attitudes, and beliefs regarding sexuality. The presence of sexual problems was assessed using the following question: 'During the last 12 months have you ever experienced any of the following for a period of 2 months or more when you: (1) lacked interest in having sex; (2) were unable to reach climax (experience orgasm); (3) reached climax (experienced orgasm) too quickly; (4) experienced physical pain during sex; (5) did not find sex pleasurable; (6) had trouble achieving or maintaining an erection (men only); and (7) had trouble becoming adequately lubricated (women only)?'. Respondents were permitted to answer yes to all that applied. For those indicating the presence of a specific sexual problem, the relative severity was assessed in a follow-up question: 'For each of these experiences, how often would you say this has occurred during the last 12 months? Would you say that this has occurred occasionally, sometimes, or frequently?'.

We restricted our analyses to only those respondents who had had intercourse at least once in the year prior to being interviewed. This procedure reduced our sample size to 9000 women and 11 205 men and tended to drop older respondents, who were sexually inactive more frequently. Thus, the prevalence of sexual problems was calculated by dividing the total number of self-reports for each problem by the total number of respondents, who were sexually active in the year prior to being interviewed, by gender. Country-specific data were grouped into clusters, according to geographic proximity, shared cultural backgrounds, and similar modes of data collection. Using the age distribution of the entire sample in the GSSAB for women and men separately, we age-standardized the prevalence estimates for each regional cluster.

A number of possible correlates of sexual problems were investigated. These included age, self-reported measures of general health status, current level of physical activity, self-report of a diagnosed vascular condition (including hypertension, diabetes, heart disease, high cholesterol, and having had a stroke), self-report of a diagnosis of depression, prostate disease (among men), having had a hysterectomy (among women), and whether respondents currently or formerly smoked. Respondents also reported how often they thought about sex—a proxy for their current level of sexual libido—and whether they agreed with the belief that aging reduced sexual desire and/or behavior. Other self-reported measures included educational attainment, whether respondents believed that their religion guided their sexual behavior, experience with divorce and financial problems in the 3-y period prior to being interviewed, the expected time horizon of their current relationships, the frequency of engaging in sex, whether they usually engaged in foreplay, and whether they were sexually exclusive.

We utilized logistic regression in this study. This approach produced adjusted odds ratios (ORs), which indicate the odds of reporting the particular sexual problem among those with a given characteristic (eg poor health) relative to people in a reference category (eg good health), controlling for all other factors in the regression analysis. In these analyses, the presence of a sexual problem included only those respondents who reported 'sometimes' or 'frequently' having the problem (ie those who indicated 'occasionally' were recoded to indicate no sexual problem). In order to evaluate the validity of pooling a specific country with the others in a regional cluster, we employed a series of interaction models between covariates and country dummy variables to test whether a specific country could be pooled in an analysis. Countries with covariate patterns that were significantly different from the pooled sample were dropped from the analysis (results are available upon request). Thus, we dropped the following countries: (1) Austria, the UK, France, Italy, South Africa, Algeria, Taiwan, Indonesia, Philippines, and Singapore in logistic regressions of orgasm problems; (2) Sweden, Israel, Mexico, Egypt, Taiwan, and Philippines in logistic regressions of lubrication difficulties, (3) Morocco, Korea, Malaysia, Philippines, and Singapore in logistic regressions of early ejaculation; and (4) the UK, South Africa, Algeria, Morocco, Korea, Malaysia, and Thailand in logistic regressions of erectile difficulties.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.