Back grey_arrow_rt.gif
 
 
Prevalence of Sexual Activity and
Associated Factors in Men Aged 75 to 95 Years A Cohort Study
 
 
  Annals of Internal Medicine Dec 6 2010
 
Zoe Hyde, MPH; Leon Flicker, MBBS, PhD; Graeme J. Hankey, MD; Osvaldo P. Almeida, MD, PhD; Kieran A. McCaul, MPH, PhD; S.A. Paul Chubb, PhD; and Bu B. Yeap, MBBS, PhD. From Western Australian Centre for Health and Ageing, Centre for Medical Research, Western Australian Institute for Medical Research, School of Medicine and Pharmacology, School of Psychiatry and Clinical Neurosciences, University of Western Australia, Crawley; Royal Perth Hospital, Perth; and Fremantle Hospital, Fremantle, Western Australia, Australia.
 
Abstract
 
Background: Knowledge about sexuality in elderly persons is limited, and normative data are lacking.
 
Objective: To determine the proportion of older men who are sexually active and to explore factors predictive of sexual activity.
 
Design: Population-based cohort study.
 
Setting: Community-dwelling men from Perth, Western Australia, Australia.
 
Participants: 3274 men aged 75 to 95 years.
 
Measurements: Questionnaires from 1996 to 1999, 2001 to 2004, and 2008 to 2009 assessed social and medical factors. Sex hormones were measured from 2001 to 2004. Sexual activity was assessed by questionnaire from 2008 to 2009.
 
Results: A total of 2783 men (85.0%) provided data on sexual activity. Sex was considered at least somewhat important by 48.8% (95% CI, 47.0% to 50.6%), and 30.8% (CI, 29.1% to 32.5%) had had at least 1 sexual encounter in the past 12 months. Of the latter, 56.5% were satisfied with the frequency of activity, whereas 43.0% had sex less often than preferred. In cross-sectional analyses, increasing age, partner's lack of interest, partner's physical limitations, osteoporosis, prostate cancer, diabetes, antidepressant use, and β-blocker use were independently associated with reduced odds of sexual activity. Living with a partner and having a non–English-speaking background were associated with increased odds. In longitudinal analyses, higher testosterone levels were associated with increased odds of being sexually active. Other factors were similar to the cross-sectional model.
 
Limitations: Response bias may have influenced findings because sexuality can be a sensitive topic. Attrition may have resulted in a healthier-than-average sample of older men.
 
Men's health problems were associated with reduced likelihood of being sexually active, suggesting that these problems are the main reason that older men cease sexual activity. Potentially modifiable risk factors include diabetes, depression, and medication use. Endogenous testosterone levels seem to predict sexual activity after adjustment for social factors and medical comorbid conditions, but it is unclear what role testosterone therapy may have in improving sexual function in older men. Given the difficulty establishing causality in observational studies, further research is required to explore this question.
 
The number of men who considered sex at least somewhat important was greater than the number who were sexually active in all age groups, suggesting that a substantial proportion of older men may have unmet sexual needs. Addressing risk factors, such as depression, diabetes, and medication use may improve sexual activity but will not benefit partnerless men. These findings highlight the need for further multidisciplinary research to explore how older men can achieve sexual fulfillment when they are widowed or when their partners are incapable of or uninterested in sex.
 
Consistent with previous studies, we found the proportion of men who were sexually active to decrease with advancing age. However, older persons tend to hold more conservative attitudes toward sex than their younger counterparts (22), so this finding may reflect some degree of cohort and period bias. Only longitudinal studies can establish the true age effect. The cross-sectional and longitudinal associations between diabetes and sexual inactivity are unsurprising. The link between diabetes and sexual dysfunction is well established and probably reflects neuropathy, endothelial dysfunction, and decreased levels of nitric oxide, which is essential to erectile response (23). The relatively strong associations observed for antidepressant use are also consistent with the literature. Impaired libido is a common sequela of depression and is thought to reflect changes in limbic neurotransmitters (23, 24). Although antidepressant medication can alleviate depression, it may worsen sexual dysfunction. In particular, selective serotonin reuptake inhibitors are often associated with decreased libido and anorgasmia (24). Similarly, the cross-sectional association between β-blockers and sexual inactivity could also reflect underlying morbidity or may be largely attributable to these agents (25). Given that independent associations were not observed for other antihypertensive agents, the latter seems most likely. Although β-blockers may cause sexual dysfunction, older patients may not voice concerns about sexual side effects with clinicians. A perception that sex is less important to older persons may lead to patients feeling that their concerns will be trivialized or that sexual side effects are unavoidable. Given that a substantial proportion of older men want to remain sexually active as they age, clinicians should counsel patients about potential sexual side effects of medications and suggest alternative agents where appropriate.
 
Sexual Activity
 
Most men (85.0%; n = 2783) provided data on sexual activity (Table 2). Of these, 857 (30.8% [95% CI, 29.1% to 32.5%]) had sex at least once in the past 12 months. Slightly more men (89.5%; n = 2930) reported the importance they attached to sex. Almost half (48.8% [CI, 47.0% to 50.6%]) described sex as at least a somewhat important part of their life. Of the 857 men who reported they were sexually active, 95.7% (n = 820) described how satisfied they were with the frequency of sexual activity. Most (56.5%; n = 463) were satisfied, whereas 43.0% (n = 353) had sex less often than liked. Four men (0.5%) reported having sex more often than preferred. The number of sexual partners in the past 5 years was reported by 2602 men (79.5%); the median response was 1 partner (interquartile range, 0 to 1; range, 0 to 40).
 

Sexual Activity, by Age
 
As Figure 2 shows, the proportion of men engaging in sexual activity decreased with advancing age (Z = -9.50; P < 0.001). Among men aged 75 to 79 years, 39.6% were sexually active (CI, 36.7% to 42.4%), whereas only 11.0% (CI, 4.2% to 17.8%) aged 90 to 95 years had sex in the past 12 months.
 
Figure 3 shows that the importance men attached to sex was also lower with increasing age (Z = -11.38; P < 0.001). Sex was described as at least somewhat important by 59.0% (CI, 56.2% to 61.8%) of men aged 75 to 79 years but by only 20.9% (CI, 12.5% to 29.3%) of men aged 90 to 95 years.
 

INTRODUCTION
 
Since antiquity, sexual activity has largely been considered the purview of the young. Classical depictions of sexuality invoke notions of vigor, youth, and fertility. Venus, the Roman goddess of love, is portrayed as young and nubile, whereas her lover, Mars, is depicted as muscular and virile. In the following centuries, sexuality increasingly fell within the religious sphere, and reproduction was deemed the chief purpose of sexual activity. Beyond the reproductive years, sex was regarded as not only inappropriate but also immoral (1).
 
The conceptualization of older persons as asexual persisted well into the 20th century. Alfred Kinsey's pioneering studies were among the first to challenge this perception, noting that some persons remained sexually active into old age (2). However, the belief that sexuality is not a concern of older persons remains entrenched, and they are often overlooked in sexual health research (3). A recent government report of the health of older Australians made no mention of sexual health (4), and the Australian Longitudinal Study of Health and Relationships, which aims to “document the natural history of the sexual and reproductive health of the Australian adult population,” did not sample adults older than 64 years (5).
 
Sexuality is an important component of overall well-being for both young and older adults (6). Although attitudes persist that older persons are not capable of or lack interest in sex (7), many consider sexuality important and wish to remain sexually active as they age (8, 9). However, epidemiologic data about sexual activity in this age group are scarce, and it is unclear what changes can be expected as part of the aging process.
 
Sexual activity has been demonstrated to decrease with age in both cross-sectional and longitudinal studies (10, 11). These changes parallel the slow but steady decline in androgen levels in men, which peak in early adulthood and decline by 1% to 2% per year thereafter (12). Although a slowly growing body of literature describes the sexual activity and behavior of older persons, normative data are lacking. Many studies comprise persons with sexual dysfunction or are hampered by small sample sizes or large age ranges, and few have focused on persons older than 70 years (9, 13, 14). We therefore designed our study to explore sexuality in a population-based cohort of 3274 community-dwelling men aged 75 to 95 years. We aimed to determine the proportion of older men who are sexually active and explore the social, physical, and hormonal factors predictive of sexual activity in this age group. We hypothesized that psychosocial factors and medical comorbid conditions would be major determinants of sexual activity but that testosterone levels would remain a predictor after adjustment for these factors.
 
Results
 
Most participants (75.5%) were married or in a marriage-like relationship, and most (74.3%) lived with a partner (Table 1). Medical conditions that could interfere with sexual function were relatively common; prostate cancer had been diagnosed in 15.7%, and 31.7% of participants had undergone at least a partial prostatectomy. Lifetime sexual behavior was reported by 85.9% of men (n = 2811). Most men reported sex only with women (96.5%; n = 2712), 1.3% (n = 35) with men and women, and 0.5% (n = 12) with men only; 1.9% (n = 52) reported that they had never been sexually active with anyone.
 
Sexual Activity
 
Most men (85.0%; n = 2783) provided data on sexual activity (Table 2). Of these, 857 (30.8% [95% CI, 29.1% to 32.5%]) had sex at least once in the past 12 months. Slightly more men (89.5%; n = 2930) reported the importance they attached to sex. Almost half (48.8% [CI, 47.0% to 50.6%]) described sex as at least a somewhat important part of their life. Of the 857 men who reported they were sexually active, 95.7% (n = 820) described how satisfied they were with the frequency of sexual activity. Most (56.5%; n = 463) were satisfied, whereas 43.0% (n = 353) had sex less often than liked. Four men (0.5%) reported having sex more often than preferred. The number of sexual partners in the past 5 years was reported by 2602 men (79.5%); the median response was 1 partner (interquartile range, 0 to 1; range, 0 to 40).
 
The 1926 men who had not engaged in sexual activity were prompted to state a reason. More than 1 reason could be given. Lack of interest was reported by 40.5% (n = 780), physical problems or limitations by 48.4% (n = 932), lack of a partner by 20.9% (n = 402), grieving by 4.1% (n = 78), and concern that children or family members would not approve by 2.3% (n = 44). Lack of interest by the partner or physical problems or limitations of the partner were cited by 39.5% (n = 761) and 22.9% (n = 441), respectively. Other reasons were given by 5.6% (n = 108); the most common were age (n = 68), medication use (n = 13), and celibacy for religious reasons (n = 7).
 
Cross-sectional Associations With Sexual Activity
 
Univariate logistic regression was performed to determine factors associated with sexual activity in the past 12 months. Associations were observed for a large number of variables, which were then entered into a multivariate model (Table 3). After adjustment, increasing age, lack of interest in sex by the partner, physical limitations of the partner, osteoporosis, prostate cancer, diabetes, antidepressant use, and β-blocker use were associated with decreased odds of sexual activity in the past 12 months. Living with a partner and having a non–English-speaking background were associated with increased odds.
 

Longitudinal Associations With Sexual Activity
 
Longitudinal factors associated with sexual activity were explored in men with sex hormone data that were collected at wave 2. As shown in Table 4, several factors were associated with sexual activity in univariate analyses. After adjustment, increasing age, lack of interest in sex by the partner, physical limitations of the partner, diabetes, and antidepressant use were associated with reduced odds of sexual activity. Living with a partner, having a non–English-speaking background, and having higher free testosterone levels were associated with increased odds. For each 1-SD increase in free testosterone levels, the odds of having had sex in the past 12 months increased by 20%. Total testosterone level was also independently associated with sexual activity, with an effect size similar to that of free testosterone levels (data not shown), but the latter seemed to be the better predictor and, for this reason, was selected in the final model.
 
Discussion
 
In this study of men aged 75 to 95 years, a substantial proportion were sexually active and considered sex to be an important part of their life. The older men were, the less likely they were to be sexually active, but sex remained at least somewhat important to one fifth of men aged 90 to 95 years, refuting the stereotype of the asexual older person. Of those who were sexually active, more than 40% were dissatisfied with the frequency of sexual activity, preferring sex more frequently. Having a partner who is capable of and interested in sex is an important determinant of whether older men are sexually active, but men's health problems seem to be the primary reason for ceasing sexual activity.
 
Consistent with previous studies, we found the proportion of men who were sexually active to decrease with advancing age. However, older persons tend to hold more conservative attitudes toward sex than their younger counterparts (22), so this finding may reflect some degree of cohort and period bias. Only longitudinal studies can establish the true age effect. The cross-sectional and longitudinal associations between diabetes and sexual inactivity are unsurprising. The link between diabetes and sexual dysfunction is well established and probably reflects neuropathy, endothelial dysfunction, and decreased levels of nitric oxide, which is essential to erectile response (23). The relatively strong associations observed for antidepressant use are also consistent with the literature. Impaired libido is a common sequela of depression and is thought to reflect changes in limbic neurotransmitters (23, 24). Although antidepressant medication can alleviate depression, it may worsen sexual dysfunction. In particular, selective serotonin reuptake inhibitors are often associated with decreased libido and anorgasmia (24). Similarly, the cross-sectional association between β-blockers and sexual inactivity could also reflect underlying morbidity or may be largely attributable to these agents (25). Given that independent associations were not observed for other antihypertensive agents, the latter seems most likely. Although β-blockers may cause sexual dysfunction, older patients may not voice concerns about sexual side effects with clinicians. A perception that sex is less important to older persons may lead to patients feeling that their concerns will be trivialized or that sexual side effects are unavoidable. Given that a substantial proportion of older men want to remain sexually active as they age, clinicians should counsel patients about potential sexual side effects of medications and suggest alternative agents where appropriate.
 
Of interest, prostatectomy was not independently associated with decreased odds of sexual activity in either the cross-sectional or longitudinal multivariate model, perhaps reflecting improvements in surgical technique. The negative association between prostate cancer and sexual activity could reflect a more invasive surgical procedure but most likely reflects the effect of hormonal intervention, which is well known to adversely affect libido and erectile function (26). An important and novel finding of our study was the moderate longitudinal association between free testosterone levels and subsequent sexual activity. Endogenous testosterone levels seem to be an independent predictor of sexual activity in men, presumably mediated through libido (27). Nevertheless, the role of testosterone therapy in the treatment of sexual dysfunction is unclear (28). Although desire and erectile function have been shown to improve in hypogonadal men who receive testosterone, men with testosterone levels in the normal range obtain little benefit (29, 30). However, many trials of testosterone therapy have not been sufficiently powered to detect small to moderate effects. Large-scale studies are needed to determine whether interventions that increase testosterone levels in older men can improve sexual interest and activity.
 
Our findings are similar to previous studies of sexuality in middle-aged and older men, which we identified through a MEDLINE search of studies published in English (keywords included sex, sexuality, sexual activity, elderly, older, geriatric, and aging). Lindau and colleagues (10) conducted a cross-sectional study of 3005 community-dwelling adults (including 1455 men) from the United States. Sexual activity in the past 12 months was reported by 84%, 67%, and 39% of men aged 57 to 64 years, 65 to 74 years, and 75 to 85 years, respectively. Among men, diabetes was associated with lack of sexual desire and with erectile dysfunction but was not associated with sexual activity. Araujo and coworkers (11) studied 1085 men from the population-based Massachusetts Male Aging Study. Participants were aged 40 to 70 years at baseline and received follow-up 7 to 10 years later. The number of times per month that the participants had sex decreased by less than 1 time for those aged 40 to 49 years, 2 times for those aged 50 to 59 years, and 3 times for men aged 60 to 70 years. Measures of sexual desire and satisfaction also declined with age. Potential mechanisms underlying these changes were not explored by the investigators. Beckman and colleagues (31) reported the results of a series of cross-sectional, population-based surveys of 1506 adults (of whom 560 were men) aged 70 years from Gothenburg, Sweden. Men were sampled from 1971 to 1972, 1976 to 1977, and 2000 to 2001. Sexual intercourse in the past 12 months was reported by 47%, 48%, and 66% of men in the first, second, and third surveys, respectively. Of those not sexually active in the final survey, 64% reported this was because of their lack of desire or capability, whereas 30% cited their partner's lack of desire or capability. The remaining 6% did not have a partner. The researchers concluded that the male partner was largely the determining factor in whether an older heterosexual couple continues to be sexually active.
 
Strengths of our study include large sample size; composition of randomly selected, community-dwelling men; and comprehensive assessment of medical comorbid conditions. In particular, use of the Western Australian Data Linkage System enhanced detection of medical conditions and surgical procedures that might affect sexual function, such as prostate cancer and prostatectomy. Limitations may include recall and response bias, survivorship effects, possible underrepresentation of men from low socioeconomic backgrounds, and cultural and linguistically diverse groups. Random sampling strategies are likely to underrepresent such hard-to-reach persons unless oversampling is used (32). We did not have the opportunity to collect serial blood samples to assess hormone levels over time. However, blood sampling at a single time point offers a reasonable estimate of prevailing testosterone levels (33). The questions assessing sexual behavior were simple and did not require detailed answers, in hopes of minimizing recall bias. Although analysts were blinded to participants' identities, our study was not anonymous, and this may have encouraged response bias. Sexuality is a sensitive topic for many and is subject to various social, religious, and legal norms. Participants may therefore be reluctant to report “socially censured” behaviors (32). We also asked participants about sexual activity and satisfaction with sex in terms of frequency, which may not correlate with measures of quality. Given attrition over successive waves of the study, it is also likely that our results reflect some degree of survivorship bias. Rates of chronic medical conditions are similar to those reported in other studies of elderly persons, but our results are probably best considered representative of relatively healthy, community-dwelling men, rather than elderly men generally. We also cannot exclude the possibility of nonresponse bias, in which men with less liberal sexual attitudes were perhaps less likely to answer the sexual activity questions.
 
The number of men who considered sex at least somewhat important was greater than the number who were sexually active in all age groups, suggesting that a substantial proportion of older men may have unmet sexual needs. Addressing risk factors, such as depression, diabetes, and medication use may improve sexual activity but will not benefit partnerless men. These findings highlight the need for further multidisciplinary research to explore how older men can achieve sexual fulfillment when they are widowed or when their partners are incapable of or uninterested in sex.
 
In conclusion, although older men are less likely than their younger counterparts to be sexually active, sex remains important to many older men, even in the 10th decade of life. Men's health problems were associated with reduced likelihood of being sexually active, suggesting that these problems are the main reason that older men cease sexual activity. Potentially modifiable risk factors include diabetes, depression, and medication use. Endogenous testosterone levels seem to predict sexual activity after adjustment for social factors and medical comorbid conditions, but it is unclear what role testosterone therapy may have in improving sexual function in older men. Given the difficulty establishing causality in observational studies, further research is required to explore this question.
 
 
 
 
  icon paper stack View Older Articles   Back to Top   www.natap.org