Testosterone is closely tied with sexual function, so the majority of publications in the immediate period after the publication of the IOM report dealt with the intersection of testosterone and sexual function. Several manuscripts investigated the validity of new patient reported survey instruments in aging men to see if they were truly reflective of the association between aging and androgens. Others continued to investigate the “aging versus disease state” conundrum: whether sexual dysfunction is a result of abnormally low testosterone levels or just another physical aspect of aging. Finally, a large meta-analysis of 17 studies published between 1979 and 2003/2004 outlined the positive and negative effects of TRT on sexual function.

In a study of two commonly used patient reported outcome (PRO) instruments, Basar et al. investigated the relationship between testosterone levels and symptoms using the Aging Male Symptoms (AMS) score and International Index of Erectile Function (IIEF) [4]. The AMS and IIEF questionnaires were administered to 348 men (21–67 years; mean 49.6 years); serum sex hormone levels (total testosterone (TT), free testosterone (FT), estradiol (E2), and dehydroepiandrosterone-sulfate (DHEA-S)) were drawn. Serum DHEA-S and FT levels and age correlated significantly with the IIEF scores; although serum total testosterone, FT, and DHEA-S levels correlated significantly with the andrologic domain of AMS, the only correlation of the total AMS score was with age.

The Androgen Deficiency of Aging Men (ADAM) questionnaire, another popular PRO, was also evaluated for efficacy in diagnosing hypogonadism. Blumel et al. administered the ADAM questionnaire and drew serum sex hormone profiles on 96 men aged 40 years and older in a cross-sectional study in Chile [5]. Total testosterone, sex hormone binding globulin (SHBG), and albumin were measured, and bioavailable testosterone was calculated. Standard scoring for the ADAM questionnaire was used to evaluate for androgen deficiency (if items 1 or 7 or any three other questions of the ADAM questionnaire were positive). A total of 78 men (81.3%) were identified as androgen deficient by the ADAM questionnaire; however, available testosterone confirmed the diagnosis in only 27 cases (28.1%). Low libido was a better predictor of hypogonadism by itself, rather than the ADAM questionnaire (63.3% sensitivity and 66.7% specificity). For now we do not have a statistically valid, rigorous PRO for androgen deficiency in aging men, and the need remains, a topic discussed in a later chapter of this text.

While efforts to find a good PRO continue, others tried to determine whether there exists a threshold at which patients become symptomatic, or whether each symptom may have an individual serum testosterone threshold at which it emerges. In such an attempt, Lackner et al. evaluated 675 healthy men using the Aging Male Symptoms (AMS) scale, the Beck Depression Index (BDI) and the International Index of Erectile Function (IIEF), and attempted to terrelate scores with AM serum testosterone levels [6]. The patients were divided into two groups: those with symptoms and those without. Patients suffering from lack of concentration, decreased libido, listlessness, and both somatic/psychological symptoms by AMS demonstrated testosterone levels that were different from those in men without these symptoms. However, this association was lost when multivariate analysis was applied; loss of libido, lack of vigor, and sexual dysfunction were associated with age rather than with testosterone. Thus, levels of serum testosterone at which specific symptoms emerged could not be identified in this study.

Two manuscripts examined the relationship of testosterone levels with erection and ejaculation. Gades et al. evaluated the association between sex hormone serum levels, erectile function, and sexual drive in a population-based sample of men (the Olmsted County Study of Urinary Symptoms and Health Status) [7]. In a random sample of men residing in Olmsted County, MN, 414 men completed the Brief Male Sexual Function Inventory (BMSFI), underwent physical examination, and had serum hormone measurements performed. Of these men, 294 had a regular sexual partner and androgen measurements after 14 years of follow-up. At 14 years, total testosterone and erectile function (but not libido) demonstrated significant correlation, despite adjustment for age. Bioavailable testosterone was significantly correlated with both erectile function and libido.

Corona et al. evaluated the role of testosterone and hypogonadism in the control of the ejaculatory reflex, comparing subjects with premature ejaculation (PE) or delayed ejaculation (DE) to those without ejaculatory dysfunction [8]. Serum hormonal and biochemical parameters were studied in 2,437 men with sexual dysfunction. Premature ejaculation was reported by 714 (25.9%) and delayed ejaculation by 121 (4.4%). Testosterone levels significantly correlated with the type of ejaculation. The youngest men (25–40 years) with PE had higher TT and FT levels; the oldest (55–70 years), with DE, had lower TT and FT levels. As would be expected from these results, hypogonadism was more common in men with DE (26%) and least common in men suffering PE (12%). Adjustment for age and libido did not change the results.

Positive correlations between threshold testosterone levels and specific sexual symptoms were found in a study of TRT reported by Seftel et al. [9]. Four hundred and six aging symptomatic hypogonadal men (mean age 58 years) were randomized to transdermal testosterone gel (50 and 100 mg/day), transdermal testosterone patch, or placebo. Patients were evaluated at 30 and 90 days after initiation of treatment for significant change in the frequency of intercourse, nighttime erections, and libido. At day 30, a significant increase was seen for all three primary outcome measures for those on 100 mg/day T gel compared to the others; the results at 90 days for sexual desire and nighttime erections vs. placebo were similar. The authors defined a “threshold average daily serum T level for sexual response” which was the testosterone level at which (1) while the testosterone level might be within the normal range, the sexual response was no different from that of the group of subjects with the lowest serum T level (0–300 ng/dL); and (2) there was a significant change when treated with testosterone compared with that of the group of subjects with the lowest serum level. This threshold f was 400 ng/dL for nighttime erections; at or below this serum level, the frequency of nighttime erections was no different than that for hypogonadal men. For frequency of sexual intercourse, the “threshold average daily serum testosterone level was 500 ng/dL, and it was 600 ng/dL for sexual desire. While the authors felt these data demonstrated a clear relationship between restoring serum T concentrations and improvement in certain parameters of sexual function, they do state that the study was not designed to explicitly measure a threshold serum level.

Isidori et al. performed a systematic review and meta-analysis of placebo-controlled studies published in the past 30 year period (1975–2004) [10]. MEDLINE, the Cochrane Library, EMBASE, and Current Contents were reviewed, of which 17 randomized placebo-controlled trials were found to be eligible. The total number of evaluable patients over the 17 studies was 656, of which 284 were randomized to testosterone and 284 to placebo (P), and 88 were treated in crossover. The majority of the patients received parenteral testosterone, though a few men received transdermal or buccal testosterone. The average length of therapy across the studies was 3 months (range 1–36 months). Results of the meta-analysis revealed that in hypogonadal men with an initial testosterone level below 12 ng/dL, TRT led to an improvement in multiple parameters: nocturnal erections, sexual thoughts and motivation, number of successful intercourses, scores for erectile function, and overall sexual satisfaction. Testosterone replacement had no effect on sexual function in eugonadal men and was equivalent to placebo. The effects of T on erectile function, but not libido, were inversely related to the mean baseline T concentration. Results of this meta-analysis must be viewed critically as the majority of the studies were performed during a period when sexual function, and in particular erectile function, was the most common outcome used to judge clinical response to testosterone replacement, and the physiology of erection and pathophysiology of erectile dysfunction were poorly understood. Conclusions from studies performed during this period are not necessarily valid given our understanding of androgens and erectile function today.

As men get older, they experience many conditions, often together, that eventually result in the inability to perform many activities of daily living. Some men may experience an increased propensity to fall and decreased independence. Elderly men also have increased incidence of anemia, higher rates of metabolic syndrome, decreased sexual function, and memory impairment. These conditions likely have multiple causes, but one likely contributing cause is a low serum testosterone concentration. When young hypogonadal men are treated with testosterone, they experience improvements in sexual function, muscle mass and strength, bone mineral density (BMD), and sense of well-being. In aging men, the benefits of testosterone therapy are controversial, particularly those related to psychosocial improvement. A limited number of studies have examined these issues since the IOM report, the following are the most important.

The degree to which androgen deficiency leads to a decline in the quality of life in elderly is unknown. To examine this question, Finas et al. investigated 24 hypogonadal men aged 50 years or older (defined as free testosterone <200 ng/dL) and compared them to 24 eugonadal age-matched controls. All of the men were under treatment for benign prostatic hyperplasia. The Short Form (SF)-12 Health Survey (physical health and mental health index) assessed health-related “Quality of Life.” The investigators modified the questionnaire to broaden it, importing “vitality” and “psychological well-being” scales from SF-36. The SF-12 physical health index was reduced in the hypogonadal group, but the mental health index was not. Patients with low FT demonstrated lower scores on vitality than eugonadal men, but no differences were detected in psychological well-being between hypogonadal and eugonadal men [11].

T’Sjoen et al. used the AMS rating scale to assess the relationship of male aging symptoms and androgen levels in asymptomatic elderly men in generally good health [12]. After the AMS questionnaire was completed, serum sex steroids levels, sex hormone binding globulin, gonadotropins, and physiologic responses were measured in 161 ambulatory, elderly men, aged 74–89. Mild psychological and mild to moderate somato-vegetative symptoms were associated with diminished serum testosterone levels as assessed by AMS. However, none of the three AMS domain scale scores significantly correlated with testosterone, free testosterone or bioavailable testosterone. In healthy elderly men, the AMS was not useful in predicting androgen levels because no consistent relationship was demonstrated.

Androgens may play a role in maintenance of cognitive function in men. Patients suffering from derangements of cognitive function such as Alzheimer’s disease may benefit from TRT. Previously, Hogervorst et al. found lower levels of testosterone in men with Alzheimer’s disease compared with controls [13]. To determine whether abnormal pituitary regulation of androgens was responsible for the testosterone deficiency, the investigators compared sex hormones (follicle stimulating hormone (FSH) and luteinizing hormone (LH), and sex hormone binding globulin (SHBG)) and testosterone in 45 men with Alzheimer’s, 15 men with other types of dementia, and 133 elderly controls. There were no observed differences in LH, FSH, or SHBG levels between Alzheimer’s disease patients and controls. However, testosterone levels were significantly lower in the men with Alzheimer’s.

The association of depression and low testosterone in men is recognized; what is not as well known is whether there is a causal relationship between low testosterone and depression. In a large cross-sectional study, Almeida et al. sought to determine whether the association between serum testosterone concentration and mood in older men is independent of physical comorbidity [14]. The authors investigated 3,987 men aged 71–89 years from a community-based setting. Patients were evaluated using the Geriatric Depression Scale (GDS-15) to assess mood; physical health was assessed with Physical Component Summary score of the 36-Item Short Form Health Survey. Two hundred and three men (5.1%) had depression. These men had significantly lower total and free testosterone concentrations than nondepressed men. Depressed men also had higher obesity levels and more physical limitations. After adjusting for these factors and for age, the association between depression and low total and free testosterone concentrations did not change.

Giltay et al. examined the reversibility of hypogonadal induced depression secondary to testosterone deficiency in a randomized, placebo-controlled, double-blind, phase III trial (ClinicalTrials.gov identifier: NCT00696748) [15]. In this trial, 184 men with metabolic syndrome and low testosterone were randomized 2:1 to receive testosterone or placebo injections. Mood, well-being, and sexual function were assessed at three time periods (baseline, 18 and 30 weeks) using the Beck Depression Inventory (BDI-IA), Aging Males’ Symptoms (AMS) scale, and International Index of Erectile Function 5-item (IIEF-5) scale. Restoration to a eugonadal state led to significant improvements in all three parameters (BDI-IA, AMS, and IIEF); the greatest improvement occurred in men with a baseline total testosterone level <7.7 ng/dL.

The majority of studies published since the IOM report have dealt with the impact of androgen deficiency and replacement on frailty in the aging, and the disability and loss of vitality that accompany frailty. These reports are comprehensive and come from large populations worldwide. The outcomes are tangible and more easily measured (i.e., bone mineral density, muscle strength, etc.). There is also more certainty that therapeutic interventions are causal to the observed improvements (Table 1.2).