|Home | About | Journals | Submit | Contact Us | Français|
In humans the circulating concentrations of dehydroepiandrosterone (DHEA) and DHEA sulfate (DHEAS) decrease markedly during aging, and have been implicated in age-associated cognitive decline. This has led to the hypothesis that DHEA supplementation during aging may improve memory. In rodents, a cognitive anti-aging effect of DHEA and DHEAS has been observed but it is unclear whether this effect is mediated indirectly through conversion of these steroids to estradiol. Moreover, despite the demonstration of correlations between endogenous DHEA concentrations and cognitive ability in certain human patient populations, such correlations have yet to be convincingly demonstrated during normal human aging. This review highlights important differences between rodents and primates in terms of their circulating DHEA and DHEAS concentrations, and suggests that age-related changes within the human DHEA metabolic pathway may contribute to the relative inefficacy of DHEA replacement therapies in humans. The review also highlights the value of using nonhuman primates as a pragmatic animal model for testing the therapeutic potential of DHEA for age-associate cognitive decline in humans.
Dehydroepiandrosterone (DHEA) and its ester, DHEA sulfate (DHEAS; together, referred to hereon as DHEA/S), are together the most abundant circulating hormones in young adult humans and nonhuman primates. Although their exact physiological function is still unclear, they represent a major source of active androgens and estrogens when metabolized in central nervous system (CNS) and peripheral tissues. A number of observations, including a unique age-related profile of production and neuroprotective and pro-cognitive effects on cultured tissue and behaving rodents, have led many researchers to investigate DHEA/S’s role in the aging process and possible therapeutic actions in learning and memory. Despite a wealth of evidence suggesting DHEA/S supplementation can improve memory in rodent models, similar actions in healthy elderly humans has yet to be demonstrated. Nevertheless, it is plausible that hormonal replacement therapies (HRTs) comprising DHEA/S, rather than more conventional sex-steroid HRT, could provide an alternative and possibly safer approach in the treatment of aging-associated human pathologies. This paper provides a brief review of the evidence, from both rodent and human studies, arguing for and against the benefits of DHEA supplementation in the treatment of age-associated cognitive decline, and also offers possible explanations for the inconsistencies in the published literature.
DHEA/S is a prohormone secreted by the zona reticularis of the adrenal glands in a highly age-specific manner. While other adrenal hormones, such as cortisol, show a relatively steady level of secretion throughout aging, DHEA/S synthesis peaks in young adulthood and declines by up to 80% in old age (Orentreich et al. 1992; Labrie et al. 1997). Indeed, it has been suggested that this decline in the DHEA:cortisol ratio underlies some of the cognitive decline associated with aging, as DHEA/S can attenuate the deleterious effects of cortisol (van Niekerk et al. 2001; Karishma and Herbert 2002). Additionally, lower levels of DHEA and DHEAS have been associated with cognitive disorders with a higher prevalence in the elderly, such as Alzheimer’s disease (Weill-Engerer et al. 2002) and depression (Micheal et al. 2000). In men (van Niekerk et al. 2001) and healthy postmenopausal women (Davis et al. 2008), endogenous DHEAS levels are associated with better cognitive ability; however, the only similar study to date in nonhuman primates failed to find such an association (Herndon et al. 1999) and studies of the frail elderly reveal an inverse relationship between DHEAS and cognitive ability (Morrison et al. 1998, 2000). As the previous studies did not simultaneously measure cortisol levels, which are significantly higher in frail versus healthy elderly humans (Varadhan et al. 2008), such findings may be due to a concurrent rise in cortisol resulting in a decreased DHEA:cortisol ratio. While the immediate effects of DHEA/S have not yet been attributed to a specific receptor, some of its protective effects may result from its conversion to sex steroids. For example, it has been estimated that 30–50% of active sex steroids in men and 75% (100% after menopause) of active sex steroids in women are derived peripherally from DHEA/S (Labrie 1991). Thus, an 80% decline in DHEA from the adrenals may be greatly enhancing cognitive deficits due to the decline in sex steroid production from the gonads.
Healthy aging is often accompanied by a decline in cognitive ability that does not meet the criteria for dementia, termed age-associated mental impairment, or AAMI (Larrabee and Crook 1994). Included in this decline are deficits in working, spatial, and episodic memory (Verhaeghen and Salthouse 1997), which, in part, is maintained by the prefrontal cortex and hippocampus. As the age-related cellular changes in these areas can be reduced by estrogen (Hao et al. 2007; Saravia et al. 2007), the age-related loss of DHEA/S may further exacerbate the age-related loss of sex steroids from the gonads, thereby potentiating the deficit in these memory domains seen in old age. Some evidence suggests that HRT involving estrogen may attenuate cognitive decline (Sherwin 2007a; Løkkegaard et al. 2002), particularly in women at risk for developing Alzheimer’s disease (Hu et al. 2006; Yue et al. 2007). These results remain controversial, however, as other HRT replacement studies, including that of the Women’s Health Initiative, have shown null effects or negative influences of HRT on cognitive decline (Craig et al. 2005; Lethaby et al. 2008). Many postulate that these discrepancies may be due to the age of HRT initiation relative to menopause, as in both humans and rodents this variable seems to determine the direction of estrogen’s effect on cognition (Kang et al. 2004; Daniel et al. 2006; Sherwin 2007b; Bohacek et al. 2008). Regardless, estrogen therapy carries with it an increased risk for breast malignancy (Rohan et al. 2008). In contrast, DHEA/S does not exhibit the same proliferative effects on breast cancer tissue as estrogen (Labrie et al. 1998), yet can be converted into estrogen in some peripheral tissues. Consequently, it is plausible that supplementation with DHEA/S may offer some of the same benefits as estrogen replacement, but with reduced risk.
While a receptor specific to DHEA or DHEAS has not been isolated, rodent studies have observed a number of effects of the steroid that, on a cellular level, may improve memory. DHEA/S has been shown to antagonize the androgen receptor and agonize estrogen receptor β; thus, it may exert some of the same actions as estradiol (Chen et al. 2005). Also, DHEA/S may affect synaptic plasticity in the hippocampus through antagonism of the GABAA receptor (Majewska 1992), facilitating long-term potentiation via NMDA agonism (Mellon and Griffin 2002; Chen et al. 2006), and enhancing glutamate release during learning (Lhullier et al. 2004). As mentioned earlier, DHEA/S significantly protects the CNS against the effects of cortisol by attenuating its suppression of neurogenesis (Karishma and Herbert 2002). Additionally, DHEA/S has been observed to be neuroprotective against oxygen-glucose deprivation (Kaasik et al. 2001), oxidative stress (Bastianetto et al. 1999; Kumar et al. 2008), and excitotoxicity (Kimonides et al. 1998; Mao and Barger 1998), as well as effective at enhancing cell survival, proliferation, and neurogenesis (Karishma and Herbert 2002; Suzuki et al. 2004). In line with this evidence, in vivo studies have shown an anti-amnestic effect of DHEAS (Flood et al. 1988) as well as an anti-aging effect on cognition (Flood and Roberts 1988) in rodents.
Although DHEA/S supplementation studies have been performed both in vitro, using cell culture, and in vivo, using rodents, in many cases endogenous neurosteroidogenesis was not blocked. Therefore, the observed effects may have been mediated by endogenous conversion of DHEA/S to estradiol. This is certainly plausible given that estradiol supplementation can produce many similar results (DeNicola et al. 2008). The proteins and enzymes necessary for the conversion of DHEA/S to estradiol, as well as for the synthesis of DHEA/S from cholesterol (Fig. 1), are present in a region-dependent manner in the rodent brain (Zwain and Yen 1999; Hojo et al. 2003; Kohchi et al. 1998; Gottfried-Blackmore et al. 2008), suggesting that this may be a valid mechanism of action. Interestingly, adrenalectomy combined with gonadectomy has no effect on the central levels of DHEA/S in rodents (Robel et al. 1987) and suppression of adrenal activity with dexamethasone has no effect on DHEA/S levels in the nonhuman primate brain (Corpechot et al. 1981), suggesting that the hormone is indeed synthesized de novo from cholesterol in the brain. Indeed, locally produced estradiol has been shown to have significant impacts on hippocampal synaptic plasticity in rodent models (Kretz et al. 2004; Rune and Frotscher 2005; Mukai et al. 2006). It is important to note, however, that while similar neurosteroidogenesis has been suggested in primates and humans (Robel et al. 1987), this pathway has yet to be investigated in detail.
Because of the strong pro-cognitive and anti-aging effects of DHEA/S previously observed in rodents, attempts have been made to examine the efficacy of DHEA/S supplementation in elderly humans. There is general agreement that DHEA supplementation can exert beneficial effects on mood and well-being in populations with adrenal insufficiency (Arlt et al. 1999; Bloch et al. 1999; Hunt et al. 2000) and depression (Wolkowitz et al. 1999; Schmidt et al. 2005), and that it shows some benefit in a primate model of Parkinson’s disease (Bélanger et al. 2006); this suggests that DHEA supplementation might also enhance cognitive functions in the elderly. Additionally, low circulating DHEA/S levels are thought to play a role in the development of Alzheimer’s disease (Weill-Engerer et al. 2002) and in some domains of memory impairment (van Niekerk et al. 2001; Davis et al. 2008), again, supporting the hypothesis that DHEA/S supplementation may improve cognition in the elderly. So far, however, clinical studies of DHEA/S supplementation have failed to provide convincing evidence in support of this hypothesis. No studies of DHEA replacement on healthy elderly populations, either acute administration or chronic (up to 12 months) supplementation, have shown a benefit in memory with treatment (Wolf et al. 1997, 1998; Wolf and Kirschbaum 1999; Arlt et al. 2001; Grimley Evans et al. 2006; Kritz-Silverstein et al. 2008), and some have even observed a negative effect on memory (Wolf et al. 1998; Parsons et al. 2006). DHEA supplementation has also shown no benefit in the treatment of Alzheimer’s disease (Wolkowitz et al. 2003).
With the substantial evidence pointing to a cognitive benefit of DHEA/S supplementation in rodents, why are similar causal effects of DHEA/S not seen in human studies? One possible explanation may lie in the method of supplementation. DHEA is marketed as a dietary supplement in the United States, most often in doses of 25–50 mg per day, and thus this is the form and dose most often chosen for clinical studies, although doses as high as 450 mg per day have also been used (Bloch et al. 1999). In rodent studies, however, the sulfated form of DHEA (DHEAS) has been commonly used. DHEA has a half-life of only 30 min while the half-life of DHEAS can be up to 12 h (Wolf and Kirschbaum 1999), thus, rapid clearance of DHEA may explain the lack of observed effects. A difference in the adrenal glands of rodents and humans may be another cause of discrepancy: while humans synthesize a large portion of DHEA/S from the adrenal glands, rodent adrenal glands do not, and it has been proposed that most, if not all, active DHEA/S is produced locally. Similarly, rodents lack the observed age-related decrease in DHEA/S production, and thus DHEA/S “replacement” may not be as physiologically relevant in a rodent model as in a primate or human model (Wolf and Kirschbaum 1999). It should also be emphasized that the human clinical studies were performed in elderly subjects, who already had attenuated circulating DHEA/S concentrations. Therefore, the studies specifically tested whether DHEA supplementation could reverse cognitive decline. They did not, however, examine whether maintenance of “youthful” circulating DHEA/S concentrations can prevent or delay age-associated cognitive decline. It is generally assumed that the beneficial effects of estrogen on cognitive function are dependent upon when the HRT is initiated relative to the onset of menopause, and so the lack of an obvious beneficial effects of DHEA supplementation are not surprising.
To fully understand the potential for DHEA/S supplementation to yield cognitive benefits in humans it might be necessary to perform studies in nonhuman primate animal models, rather than rodents. Rhesus macaques, for example are long-lived primates that show many similar aging-associated changes in cognitive function to those seen in humans. Furthermore, like humans rhesus macaques show similar adrenal gland structure and endocrine function (Conley et al. 2004; Nguyen and Conley 2008; Abbott and Bird 2008), and show a marked age-related decline in circulating DHEAS concentrations (Urbanski et al. 2004; Downs et al. 2008). Also, like women, female rhesus macaques undergo menopause (Downs and Urbanski 2006). One possible target of investigation includes the neurosteroidogenic pathway (see Fig. 1). An age-related decline in expression of the enzymes necessary to form estradiol from DHEA/S may explain the negative results reported by clinical studies. In particular, the enzyme responsible for the age-related decline in DHEA/S production in the adrenal glands, 17,20-desmolase (Liu et al. 1990; encoded by the p450c17 gene and necessary in forming DHEA from pregnenolone), may also decline in central tissues, resulting in reduced central production of DHEA. Other genes responsible for the conversion of DHEA itself to estradiol may decline in expression as well. Targeting these enzymes instead of, or in addition to, DHEA/S supplementation may provide another route of cognitive therapy in healthy older adults. Because rhesus macaques can be maintained under tightly controlled environmental conditions (including temperature, photoperiod, diet, and medication), and subjected to long-term in vivo studies, they may represent an ideal animal model in which to comprehensively examine the therapeutic potential of DHEA in elderly humans. Moreover, because rhesus macaques can yield high quality mRNA from post-mortem tissues, in vitro studies could help to elucidate the underlying mechanisms by which DHEA/S exert its beneficial effects within the CNS (Racchi et al. 2003).
This work was supported by National Institute of Health grants: AG-019914, AG-026472, AG-029612 HD-29186, and RR-00163.