A review of studies examining fertility at different ages demonstrated significant age-related differences in men, including lower pregnancy rates, increased time to pregnancy, and subfecundity in men older than 50 years. Several studies have reported that the fertility rates decline with advancing age (383-384, 393), but some recent studies have shown no age-related changes in fertility (400). Some changes in fertility rates might be related to age-related decrease in sexual activity. A literature review found no significant change in sperm concentration with aging when comparing men under the age of 30 to those greater than 50 years (388). However, in general, semen volume, sperm motility, and the number of morphologically normal sperm decrease with advancing age (Table 7; 383-396). A number of these studies, however, did not control important confounding variables. Of the 21 studies in which sperm densities were compared among men of different age groups (388), only four studies adjusted for the duration of abstinence, well known to affect sperm concentration. In addition, there is significant heterogeneity in the populations studied; most of the studies examined data from semen of sperm donors while others examined men from infertility clinics. Sperm donors might represent a healthier group of men than the general population; conversely men in infertility clinics might be more likely to have abnormalities of sperm number or function. Even studies that have controlled for abstinence as well as alcohol and tobacco use have shown an age-related decrease in semen volume. In one study of men whose partners had bilateral tubal obstruction or absence of both tubes and who were treated by conventional IVF, the odds ratio of failure to conceive was higher for men 40 years of age or older (397). This provocative study that needs further confirmation suggests that paternal age over 40 years might be a risk factor for failure to conceive.

Necropsies on adult men of different ages have revealed that the testicular volume was lower only in men in the 8th decade of life (423). A recent study examined testicular germ cells obtained by orchidectomy from 36 older men with advanced prostate cancer and by testicular biopsy from 21 younger men with obstructive azoospermia, as controls (401). The ratios of primary spermatocytes, round spermatids, and elongated spermatids to Sertoli cells were significantly decreased in the testes of older men, but the ratio of spermatogonia to Sertoli cell number remained unchanged (401-402). The reason for the lack of change in number of spermatogonia in older men compared to younger men has been explained by a significantly lower apoptotic rate and lower cell proliferation compared with that of younger men, suggesting that germ cell proliferation and apoptosis diminish with aging (402).

Other studies evaluating the fidelity of the germ cell compartment are primarily cross-sectional and depend on analyses of sperm number and semen quality; large-scale chromosomal analyses in healthy community dwelling men are scarce as most data are derived from fertility clinics. A review of studies examining semen quality at different ages demonstrated significant age-related decreases in semen volume and sperm morphology. The change in sperm morphology has been hypothesized to be due to an increase in aneuploidy with age. Härkönen et al (391) found that sperm morphology was directly associated with the number of chromosomes in sperm and that men with higher aneuploidy rates for chromosomes 13, 18, 21, X and Y had lower sperm motility and sperm concentrations. However, a recent study in a small sample of men found no significant increase in the risk of aneuploidy for men over the age of 60 (393) In spite of the changes in sperm morphology and motility from older men, in vitro fertilizing capacity of the sperm is well preserved (396-397). In some older men, degenerating germ cells can be observed suggesting loss of germ cells with age.

There are several difficulties in interpreting these data on age-related changes in sperm density and function. The normal range for sperm concentration in men is wide where sperm concentration above 20 million/ml (total sperm per ejaculate >60 million) is considered normal. Thus, even though average sperm concentrations might decline with aging, they might still be in the normal range (> 20 million/ml) (396, 400-401). Furthermore, normal sperm counts might not always correlate with normal sperm function.

Studies in flies demonstrate more germ cells during larval than adult stages suggesting age-related quiescence of the germ line (404). Significant age-related decreases in germ cells and spermatogenesis also have been reported in rodents and primates (405-409). The Brown Norway rat has been studied as a model of aging of the human male reproductive system because in this rodent model, serum testosterone levels decrease with aging, as they do in humans (406-408). Along with changes in hypothalamic-pituitary hormones, alterations in sperm counts, sperm maturation, Sertoli cell number, and progeny outcomes have been observed in this rodent model (Table 8; 406-409, 436-442). Analysis of ribosomal DNA from germ cells of the male brown Norway rat revealed hypermethylation of ribosomal DNA (409). Alterations in ribosomes have been theorized to promote aging of cells by multiplying errors in protein synthesis which initially might elude gross morphological analysis but eventually might lead to germ cell degeneration (410). Further assessment of spermatogonial stem cell populations is needed, but altered DNA methylation of germ cells with aging could result in decreased fecundity. In many animal models of life span extension, there is a trade-off between longer life and fecundity, although there are some exceptions (411).

Since Sertoli and Leydig cells are crucial to spermatogenesis, changes in these cells could affect sperm number and function. In stallions, the numbers of Sertoli cells decreases with aging but individual Sertoli cells display a remarkable capacity to accommodate greater numbers of developing germ cells (412). In men, Sertoli cell number has been reported to be lower in men aged 50 to 85 years than in men aged 20 to 48 years (413). The apoptotic rate of primary spermatocytes in aged men was also significantly elevated compared with that of younger men, resulting in a decrease of the number of primary spermatocytes per Sertoli cell (402). These latter data suggest a failure of Sertoli cells to support spermatogenesis.

Sertoli cells produce inhibin, which regulates gonadotropin expression from the pituitary. Inhibin B has been identified as the physiologically important form of inhibin in men and has been validated as a valuable serum marker of Sertoli cell function and spermatogenesis. Higher gonadotropins and lower inhibin levels in older men suggest a decline in Sertoli cell function (413); however changes in circulating inhibin B levels with advancing age have been inconsistent (413-416). Overall, these data suggest a possible decline in Sertoli cell number and function in older men with little affect on spermatogenesis.

Aging is accompanied by a progressive, albeit variable, decline of Leydig cell function with a decrease of mean serum free (or bioavailable) testosterone levels in the population between age 25 and 75 years (417). Total Leydig cell volume and the absolute number of Leydig cells per individual decline as a function of age, although individual Leydig cell volume and total testis weight do not change with age (417-421). In one study, age accounted for more than a third of the variation in Leydig cell number, and explained more than half the variation in daily sperm production (420). This might in part be explained by a fusion of Leydig cells producing fewer but multinucleated cells with age (421). The functionality of the multinuicleated cells is not known.

In a comparison of 66 younger men (21-25 years) to 134 older men (>50) referred for andrological evaluation, after adjustment for duration of sexual abstinence, the ejaculate volume, progressive sperm motility, and sperm morphology were lower in older men (422). The older men also had a higher frequency of sperm tail defects, suggesting epididymal dysfunction (423). In addition, the fructose content was significantly lower in older men suggesting a defect in the seminal vesicle contribution to semen (423). There were no significant differences in sperm concentration and testicular size between the young and older men in this study.

Age-related changes in the supporting structures for sperm maturation have been described in the Brown Norway rat. These changes include reductions in the numbers of Leydig and Sertoli cells, seminiferous tubules, and in epididymal cell number and function (406-409). Changes in the supporting cells and structures for sperm maturation have been invoked to explain the age-related decrease in sperm number and fecundity in rats.

Although concern has been expressed about age-related increases in germ cell mutations and impairment of DNA-repair mechanisms (425-428), and the men have been blamed for introducing the majority of new mutations into the gene pool (426), the frequency of chromosomal aneuploidy or structural abnormalities is not consistently increased in sperm of older men (392-395, 425-428), The incidence of some autosomal dominant disorders, such as achondroplasia, polyposis coli, Marfan syndrome, and Apert syndrome has been reported to be increased in the offspring of men with advanced age, consistent with transmission of sporadic missense mutations (426-435). The minimum absolute frequency of autosomal dominant disease attributable to mutation in sperm of fathers over 40 years of age is 0.3 to 0.5% (432). These data suggest that approximately one third of babies with diseases due to new autosomal dominant mutations are fathered by men aged 40 years or older.

The existence of a paternal age effect on Down syndrome is controversial. Earlier studies from the 1960s and 1970s found no correlation between Down syndrome and paternal age (e.g., 430). However, a study in New York from 1983 to 1997 found a significant greater numbers of mothers and fathers 35 years of age and older, respectively among parents of patients with Down’s syndrome (431). Among the cases of Down syndrome evaluated, paternal age had a significant effect only in mothers 35 years of age or older, and was greatest in couples greater than 40 years of age where the risk was 6 times the rate of couples younger than 35 years of age (431).

Some cardiac defects have also been attributed to aberrant genetic input from older men. For instance, a case-control study of 4,110 individuals with congenital heart defects born between 1952 and 1973 in British Columbia, found a general pattern of increasing risk with increasing paternal age among cases relative to controls for ventricular septal defects, atrial septal defects and patent ductus arteriosus (432-433). The risk of schizophrenia has also been reported to increase with paternal age (434) and possible loci affecting this risk have been identified (435). In addition, a modest proportion of preeclampsia, normally associated with increased maternal risk factors including age, might be attributable to an increase in paternal age although no gene loci have been identified (436). These observations need further corroboration.