Aging is associated with important changes in body composition that may be influenced by the endocrine milieu and can have important endocrine and metabolic consequences (Table 56-1). In cross-sectional studies, body weight increases until about age 55 years and then declines. This may be the result of a “die-off” effect in the heaviest patients during middle age. Prospective studies suggest that weight actually declines after age 65 to 70 years. This reduction in body weight, whether intentional or unintentional, appears to be associated with an increase in mortality, morbidity, and disability. Indeed, there is a well-described obesity paradox in older patients, so that the weight with the lowest overall mortality is shifted upward. The explanation for this is not clear, but it is possible that any sustained weight loss may, in fact, be unintentional, given that intentional weight loss is difficult to maintain. Weight loss in the presence of illness or disease that raises cytokine levels may predispose to a disproportionate loss of weight as lean mass (muscle mass), thereby exacerbating age-related “sarcopenia” and leading to a catabolic state. It is also possible that the apparent obesity paradox of aging results from the heterogeneity of obesity in older patients; obesity beginning in young or middle age is associated with untoward consequences, and obesity beginning in older age is less dangerous. Interventions in these two potentially heterogeneous groups may also be expected to produce different effects, but this has not been studied.

There is an inevitable loss of lean body mass, mostly skeletal muscle, with aging. The aging-associated loss of muscle has been termed sarcopenia and has been blamed for much, but not all, of the age-related decline in muscle strength and power (dynopenia). In cross-sectional studies, a 20% to 30% loss of lean mass has been detected between ages 30 and 80 years. The decline in strength is even greater, with longitudinal studies finding up to a 60% loss from age 30 to 80 years. Furthermore, the loss of strength and power is not as linear as the loss of muscle mass and seems to accelerate at older ages. A 25% decline in strength has been detected between 70 and 75 years of age. Power (work per unit time) may decline at double the rate of strength. These changes in lean mass, muscle mass, strength, and power have complex but important functional consequences for older people. Of greatest clinical importance are the relationships among muscle mass, strength or power, and functional ability. These relationships are complex and probably nonlinear, and in general they have been difficult to demonstrate clearly. An example is the consistent association of testosterone supplementation with increases in lean mass but inconsistent associations with improvements in strength or function.

A more recent concept, sarcopenic obesity, considers the degree of adiposity relative to lean mass. Currently, because consensus is lacking on how sarcopenic obesity should be defined, the prevalence and clinical relevance have not been established. One contributing factor in sarcopenic obesity could be weight cycling, with loss of lean and fat mass followed by regain of fat mass only. This may be more likely in older patients who are less anabolic and generally less active. The loss of lean mass with aging can have a profound effect on resting metabolic rate and thus can predispose to further accretion of fat mass if caloric intake is not reduced.

Prospective data indicate that peak bone mass occurs during the late teen years in women and about a decade later in men. Because of the intimate structural and functional link between muscle and bone, the occurrence of peak bone mass likely corresponds to peak skeletal muscle development. It is generally thought that bone mass is maintained, or decreases slowly (< 0.2% per year), at least through age 40 years in women and age 50 years in men. Intuitively, a decline in physical activity during middle age may be expected to induce an even faster rate of bone loss. However, the increase in body weight that typically also occurs during middle age may counter this to a large extent, by increasing mechanical loading forces acting on the skeleton during weight-bearing activity. The inevitable loss of bone mass in old age increases the risk of osteoporosis in elderly men and augments the risk of osteoporosis in postmenopausal women. In elderly women and men, the decrease in bone mineral at the hip appears to be accelerated (∼1% per year) relative to the changes at the spine, which may increase in advanced age. Vertebral compression fractures and the development of extravertebral osteophytes lead to an increase in bone mineral density (BMD) that does not reflect increased vertebral bone strength. The utility of spine BMD for the diagnosis of osteoporosis in elderly persons is therefore compromised.

There is an accelerated decline in BMD around the time of menopause in women. What remains somewhat controversial is whether the menopause-induced increase in bone resorption diminishes after a few years or persists into old age. In this regard, observational studies of women 65 years old and older indicate that the rate of bone loss continues to increase with age, particularly in the hip region. This finding is corroborated by observations that serum markers of bone turnover increase at menopause and remain elevated into old age.

It is unlikely that even vigorous weight-bearing exercise can completely mitigate the deleterious effects of estrogen deficiency on BMD. Older female athletes who are not taking hormone replacement therapy have lower BMD than do premenopausal athletes. Moreover, young female athletes with menstrual cycle dysfunction can have BMD levels in the osteopenic range (1.0–2.5 SD below the average peak BMD) and even the osteoporotic range (> 2.5 SD below the average peak BMD), despite participation in sports that involve high levels of mechanical loading (e.g., gymnastics, distance running).

Although the direct effects of estrogens on bone metabolism are well known, growing evidence indicates that the responses of bone cells to mechanical stress involve activation of estrogen receptor alpha. The effects of age-related sex hormone deficiency on receptor density and/or function in bone remain unknown. In animal models, the effects of mechanical stress in the presence of estrogens (in females) or androgens (in males) on the bone proliferative response have been found to be either additive or synergistic (i.e., more than additive). There is also evidence for additive or synergistic effects of exercise and estrogens on BMD in postmenopausal women. More recent studies of laboratory animals suggest that estrogen receptor alpha may facilitate the effects of mechanical loading on bone, whereas estrogen receptor beta may inhibit such effects.

Dual-energy x-ray absorptiometry (DXA) scans are the best means of following changes in bone density in the elderly population. T-scores higher than −1.0 indicate normal bone density, whereas T-scores lower than −1.0 but higher than −2.5 indicate osteopenia, and T-scores lower than −2.5 indicate osteoporosis. Pharmacotherapy for low bone density is not only reserved for those with osteoporosis. The World Health Organization (WHO) developed the FRAX score, which gives an individual’s 10-year probability of fracture and can identify those patients with osteopenia who would benefit from treatment. The recommendation in the United States is to consider treatment if the 10-year probability of a major osteoporosis-related fracture is at least 20% or at least 3% for a hip fracture. Additionally, anyone with a fragility fracture regardless of BMD should be considered for treatment. A FRAX score may be calculated without DXA.

Following the supplementation of calcium and vitamin D, bisphosphonates are the first-line drug class for the improvement of bone density and fracture prevention. Their antiresorptive properties have been shown to be efficacious in improving BMD and preventing fractures at all major sites. There have been reports of rare side effects such as osteonecrosis of the jaw, as well as concerns about an association between long-term use (> 3–5 years) of bisphosphonates and atypical fractures of the thigh, known as subtrochanteric and diaphyseal femur fractures. Recommendations are to consider periodic reevaluation of the need for continued bisphosphonate therapy, particularly in patients who have been treated for more than 5 years, and always to ask about symptoms such as new thigh or groin pain.

Denosumab is an antiresorptive osteoporosis drug, approved by the Food and Drug Administration (FDA) in 2010, that is a human monoclonal antibody functioning as a RANK ligand (RANKL) inhibitor. This drug has the net effect of preventing the maturation of osteoclasts, decreasing bone resorption with increased BMD in women and men, and reducing the risk of fracture (better fracture data in women). Use of this agent could be considered in patients who have been unresponsive to or are intolerant of other available osteoporosis therapies and in patients with renal insufficiency. No dose adjustment is necessary in patients with renal impairment, although they may be at more risk for hypocalcemia, particularly if they are not receiving adequate calcium and vitamin D. Additionally, because RANKL functions within the immune system, long-term monitoring is needed to assess increased risk for serious infections and neoplasms.

The only FDA-approved anabolic drug is teriparatide, which is recombinant human parathyroid hormone (PTH[1–34]). It improves BMD and prevents fractures. This drug is administered via daily subcutaneous injection, an important consideration in elderly patients who may not have the functional capacity or appropriate assistance to do so.

There is an increase in total adiposity and shift toward more abdominal fat distribution with advancing age. The increase in central adiposity begins in young men who gain excess fat, but this does not appear to occur in women until around the time of the menopausal transition. Although the loss of lean mass was once thought to be the primary determinant of physical disability in old age, more recent studies indicate that increased adiposity is an independent, and perhaps stronger, predictor of disability in older individuals. The increase in abdominal visceral adiposity (along with the decline in physical activity) plays an important role in the age-associated increase in insulin resistance and probably contributes to the high incidence and prevalence of type 2 diabetes mellitus and metabolic syndrome in old age.

Cross-sectional comparisons of women across the age spectrum suggest that waist size increases more rapidly in women aged 50 years and older than in younger women. Prospective studies indicate that increases in waist circumference are related to both chronologic and ovarian age, with the most rapid increases in waist girth occurring in perimenopausal women. Premenopausal women treated with gonadotropin-releasing hormone agonists to suppress sex hormones gain 1 to 2 kg of fat mass in 4 to 6 months, with a disproportionate increase in central body regions. Several randomized, controlled trials provided evidence that postmenopausal women who took hormone therapy gained less weight and had less increase in waist size than did placebo-treated women. The effects seemed to be slightly larger with unopposed estrogens. It has not yet been determined whether estrogens specifically prevent or attenuate intraabdominal fat accumulation.

Obesity in older adults is a mounting public health concern, given its increasing incidence and its association with loss of functional independence and frailty. Hypocaloric diets have been effective in reducing total and visceral fat and improving glucose tolerance, insulin sensitivity, blood pressure, and pulmonary function. Obese elderly adults are capable of participating and adhering to rigorous interventions such as diet, exercise, or diet plus exercise. Such studies have found that diet alone and exercise alone both reduce frailty, but the combination of diet and exercise generates the greatest objective functional and subjective benefits.

Intentional weight loss typically results in a loss of lean mass (muscle and bone), which may exacerbate sarcopenia and the risk of osteoporosis. This could have adverse effects in elderly adults who are already at risk for osteoporosis. The addition of exercise training to diet in older obese adults prevented the weight-loss-induced increase in bone turnover and attenuated, but did not prevent, the decline in BMD.

Some prospective observational studies have suggested that weight loss in older adults may be associated with increased mortality, despite a decrease in comorbidities such as cardiovascular disease and type 2 diabetes. In randomized controlled weight loss interventions, weight loss did not increase mortality in older adults over 8 to 12 years of follow-up. In fact, secondary analyses from one trial suggested that intentional weight loss may reduce mortality risk in this population. Additional trials of intentional weight loss in older adults are needed to confirm whether it does, indeed, reduce mortality risk and whether the risk-to-benefit profile is similar in older adults who became obese earlier versus later in life.

Weight loss surgery is effective at reducing medical comorbidities in elderly persons. The sparse available data do not suggest increased mortality. In fact, mortality rate may be decreased compared with matched obese cohorts.

Vitamin D supplementation has been found to reduce the incidence of osteoporotic fractures in elderly persons. This may occur via increased bone mineralization and/or improved muscle function and reduction in falls. Vitamin D deficiency is defined as a 25-hydroxyvitamin D (25-OHD) level of less than 20 ng/mL (50 nmol/L). It has been estimated that more than 40% of community-dwelling older women and men in the United States are vitamin D deficient, and the prevalence is even higher in nursing home residents. There are multiple causes of vitamin D deficiency in older adults, including the following: decreased skin synthesis; decreased sun exposure; decreased intake; impaired absorption, transport, or liver hydroxylation of oral vitamin D; medications altering vitamin D metabolism; chronic illnesses associated with malabsorption; and liver and kidney disease.

BMD is adversely affected when serum 25-OHD is less than 30 ng/mL. Vitamin D₃ supplementation of 700 to 800 IU/day or 100,000 IU every 4 months has been found to raise serum 25-OHD to more than 30 ng/mL and reduce the incidence of fractures. There is currently no evidence for antifracture efficacy of vitamin D₂ supplementation.

Vitamin D deficiency also causes muscle weakness. Proximal muscle strength is linearly related with serum 25-OHD when levels are less than 30 ng/mL. Vitamin D supplementation has been associated with a 22% reduction in falls. Nursing home residents randomized to receive 800 IU/day of vitamin D₂ plus calcium had a 72% reduction in falls.

In addition to its important role in muscle and bone metabolism, vitamin D deficiency is postulated to influence immune function, cancer risk, PTH and renin production, and insulin secretion. Epidemiologic studies demonstrate higher mortality in patients with insufficient or deficient levels of 25-OHD.

The recommendations for vitamin D supplementation in older adults differs among professional societies. However, it is agreed that daily supplementation is best achieved with vitamin D3. In 2010, the National Osteoporosis Foundation (NOF) recommended that older adults should have a serum 25-OHD level of 30 ng/mL (75 nmol/L) to reduce risk for falls and fractures. Supplementation doses up to 800 to 1000 IU/day were recommended because of the lack of evidence for the efficacy of higher doses.

In 2012, the United States Preventive Services Task Force (USPSTF) reported that supplementation of 400 IU/day vitamin D in combination with 1000 mg/day calcium does not reduce fracture risk in noninstitutionalized, community-dwelling, asymptomatic adults without a previous history of fractures. It was further noted that evidence regarding the effectiveness of higher doses of vitamin D and calcium on incident fracture is lacking.

The current Institute of Medicine recommendations for vitamin D supplementation are set at 600 IU/day for men and women aged 51 to 70 years and 800 IU/day for older individuals.

Studies of yeast, worms, flies, rodents, and mammals have demonstrated that caloric restriction (CR; 30%–40% reduction in daily energy intake) increases mean (i.e., average life expectancy) and maximal life span. Generating a negative energy balance in rodents through increased energy expenditure (exercise) results in similar improvements in mean life span as CR, but it does not increase maximal life span. Long-term studies of CR in humans and other primates are under way, but short-term studies suggest that CR produces physiologic, metabolic, and hormonal effects that parallel many of the positive effects found in other species.

It is estimated that one fourth to one third of the differences in life expectancy in humans may be explained by genetic factors, but currently no definitive biomarkers or genes are associated with longevity in humans. Large-scale collaborations, such as the pan-European Genetics of Healthy Aging consortium and the United States Longevity Consortium, are studying different populations to address this issue.

Total testosterone (TT) concentrations decline with age (Table 56-2). Additionally sex hormone–binding globulin (SHBG) levels increase with age, which results in an even greater relative reduction in calculated bioavailable testosterone and free testosterone (FT) with age (declines of −14.5% for TT versus −27% FT per decade of aging).

Total plasma estradiol levels in adult men do not change significantly with age, but bioavailable and free estradiol levels decrease because of the increase in SHBG with aging (estradiol binds to SHBG with half the affinity of testosterone). In absolute terms, serum estrogen levels of elderly men are somewhat higher than those of postmenopausal women (average of 33 pg/mL versus 21 pg/mL).

In addition to an age-related decline, changes in health and lifestyle factors, such as weight gain, illness, polypharmacy, glucocorticoid use, and widowhood, can reduce TT. For instance, the decline in TT associated with becoming obese (−12%) is comparable to that associated with 10 years of aging among subjects whose obesity status is stable (−13%). Low endogenous testosterone levels are predictive of future development of metabolic syndrome, as well as cardiovascular, respiratory, and all-cause mortality and cognitive dysfunction. It is not known whether raising testosterone levels by supplementation translates into decreased mortality.

The prevalence of male hypogonadism is not known because of the lack of consensus on the definition of hypogonadism with aging. The development of a consensus definition is complicated by several factors: (1) whether there should be an age-specific testosterone reference range or whether hypogonadism should be defined in relation to young male testosterone levels; (2) whether the definition should be based on TT (SHBG bound + albumin bound + free), bioavailable testosterone (albumin bound + free), or FT levels; (3) concern regarding the reliability and variability of immunoassays versus mass spectroscopy; (4) the finding that formulas for calculating bioavailable testosterone and FT may not be valid in some or all older populations; and (5) whether hypogonadism should be defined in relation to a low serum testosterone concentration alone, a low concentration plus symptoms, or symptoms alone. When defined as TT less than 300 ng/dL and FT less than 5 ng/dL, almost 50% of men with hypogonadism who were more than 50 years old were asymptomatic, and 65% of men with symptoms had normal testosterone levels. The prevalence of symptomatic androgen deficiency is estimated to be at least 5% in men aged 50 to 70 years and 18% in older men.

Of the few randomized controlled trials conducted in healthy older men, most found an increase or maintenance of fat-free mass (bone and muscle) and a decrease in fat mass (including abdominal visceral) in response to testosterone. Whether such physiologic effects translate into strength or functional improvements remains equivocal. Anemia from androgen insufficiency improves with testosterone therapy. Improvements in sexual function and sense of well-being have also been inconsistent.

The lack of consistent findings among trials of testosterone supplementation is likely related to variability in study cohorts (e.g., baseline testosterone levels, symptoms, body composition, comorbidities, physical function), the type of testosterone supplementation therapy (e.g., oral, transdermal, intramuscular, dose, and average testosterone concentration achieved), and the duration of intervention (e.g., months versus years).

In 2010, researchers from the Testosterone in Older Men with Mobility Limitations (TOM) trial published their data on adverse events associated with testosterone administration, which led to early termination of the study. The original purpose of the study was to determine the effects of testosterone administration on lower extremity strength and physical function in older men with significant mobility limitations and low serum levels of either TT or FT. The participants were men at least 65 years old with limitations in mobility and a high prevalence of chronic disease (diabetes, hypertension, obesity). During the 6-month intervention phase, those men who were randomized to testosterone gel therapy with a goal of attaining a TT level greater than 500 ng/dL unexpectedly had a higher prevalence of cardiovascular, respiratory, and dermatologic adverse events, even after adjustment for baseline risk factors. The increased frequency of cardiovascular events led to trial termination. However, this has not been confirmed in other studies or in metaanalyses of testosterone intervention studies. In addition, a more recently completed study demonstrated a reduction in cardiovascular adverse events in a testosterone-supplemented versus a placebo-treated group of generally healthy men with low normal TT levels at baseline.

Additional adverse effects include worsening of benign prostatic hypertrophy and polycythemia, likely dose related and worsened in the setting of sleep apnea. There is consistent lack of evidence of increased prostate cancer risk in studies of men receiving testosterone therapy.

The most recent guidelines from the Endocrine Society in 2010 recommend screening for androgen deficiency in older men who have consistent signs and symptoms of low androgen levels. These guidelines recommend the use of a high-quality assay measuring the TT level in the morning and confirming a low result with a repeat TT level and/or free or bioavailable testosterone level. Even if low testosterone levels are confirmed, only men with clinically significant and symptomatic androgen deficiency should be considered for treatment. If therapy is initiated, the clinician should ensure that the patient understands the uncertainty of the risks and benefits of testosterone therapy. The choice of supplementation is left up to the discretion of the clinician and patient preference. Although the Endocrine Society advises a target TT level in the midnormal range when prescribing testosterone therapy, many clinicians aim for TT levels in the low normal range to avoid potential cardiovascular or respiratory side effects, despite a lack of evidence supporting this practice.

At this time, testosterone replacement should continue to be reserved for the minority of men with frankly low serum testosterone levels and clear clinical symptoms of hypogonadism who do not have an existing clear contraindication for androgen therapy (prostate cancer, severe obstructive uropathy, liver disease, polycythemia, untreated or poorly controlled obstructive sleep apnea, and poorly controlled heart failure).

This has been an area of controversy since the completion of the Women’s Health Initiative (WHI) trials. Similar to debates about testosterone replacement, controversy exists regarding who should be treated (age, years menopausal, symptomatic), by what formulation (conjugated estrogens versus estradiol, progesterone versus medroxyprogesterone acetate (MPA), continuous versus intermittent progestins), at what dose (fixed versus target serum estradiol level), by what route (oral, transdermal, transvaginal), and for what length of time. The WHI trials generated important results, but they raised equally important questions. Oral conjugated estrogens with or without MPA may not have the same effects (good or bad) as transdermal estradiol with or without progestins. The WHI trials appear to support the “timing hypothesis,” that cardiovascular benefits may occur when therapy is initiated near the time of menopause. However, initiating hormone treatment after 10 or more years of estrogen deficiency may increase the risk of cardiovascular events.

The loss of estrogen with menopause appears to be linked to deleterious changes in body composition, including increased central fat accumulation and decreased BMD, which translates long term into increased risk for cardiovascular disease and fractures. Additionally, the loss of estrogen is associated with hot flashes, decreased sleep quality, vaginal dryness, and worsening of mood disturbances, the sum of which equals a decreased quality of life for many women.

Currently, estrogen therapy is indicated for relief of menopausal symptoms that are not relieved by other methods, with the lowest dose used for the shortest time possible. Transdermal estradiol appears to be associated with fewer thromboembolic events than oral estrogens. Because continuous conjugated estrogens plus MPA were associated with an increased incidence of invasive breast cancer in the WHI (whereas conjugated estrogens alone were not), intermittent progesterone may be a better alternative for endometrial protection.

Dehydroepiandrosterone (DHEA) and its sulfate (DHEAS), collectively referred to as DHEA/S, are the most abundant steroid hormones in humans, with approximately 95% coming from the adrenal glands. DHEAS is one of the best biologic markers of human aging. Peak serum DHEAS levels are reached early in the third decade and then decline steadily. By age 60 to 70 years, circulating levels are only about 20% of peak levels. The decrease in DHEA/S with aging does not represent a general decline in adrenal function because similar changes in other adrenal hormones do not occur.

Despite the abundance of DHEA/S and the distinctive age-related changes, little is known about the biologic effects of DHEA/S in humans. The actions of DHEA/S in humans are thought to be mediated primarily through conversion to sex hormones, and thus it may function as a large storage pool of prehormone. DHEA is the precursor for 30% to 50% of androgens in older men and for more than 70% of androgens in older women, and it is a major source of estrogens in men and postmenopausal women. The decline in DHEA/S with aging may contribute to physiologic changes that occur as a result of sex hormone deficiency (e.g., the loss of bone and muscle mass). Other proposed biologic effects include increased insulin-like growth factor-I (IGF-I), antiglucocorticoid effects, and antiinflammatory effects via peroxisome proliferator-activated receptor-alpha (PPARα) agonism.

In the United States, DHEA is considered a dietary supplement and therefore is not an FDA-regulated drug. Hence, over-the-counter products vary greatly in the amounts of bioactive hormone they contain (if any) and may have quite different pharmacokinetic profiles. Even batch-to-batch variability within a brand can be great. Despite being labeled a “dietary supplement,” DHEA has measurable effects on concentrations of hormones. In older adults, DHEA, 50 mg of bioactive hormone per day, results in the following: 300% to 600% increases in plasma DHEAS concentration in men and women; a 100% increase in plasma testosterone in women, with nonsignificant changes in men; a 70% to 300% rise in plasma estradiol in women and a 30% to 200% increase in men; and increases in IGF-I of 25% to 30% in women and 5% to 10% in men. However, the physiologic effects of DHEA supplementation in humans appears quite variable.

In randomized, placebo-controlled trials of 1 to 2 years, DHEA replacement alone in older adults did not result in significant changes in fat or muscle mass or metabolic improvements. Studies of DHEA plus an exercise stimulus (endurance, resistance, or both) have shown mixed effects. In postmenopausal women, 12 weeks of DHEA was not more effective than placebo at potentiating effects of endurance and resistance exercise on body composition, glucose, and lipid metabolism. In contrast, 16 weeks of DHEA improved muscle volume and strength tests compared with placebo when combined with high-intensity resistance exercise in older women and men.

Studies of DHEA on BMD have shown trends for increases in indices at the hip, but improvements at other sites appear to be more study and sex specific. None of the studies has been powered to demonstrate antifracture efficacy. The increases in BMD in response to short-term DHEA replacement therapy have generally been small (1%–2%). However, one study that looked at 1- to 2-year therapy with DHEA on BMD investigated whether restoration of DHEA levels to those of young adults would be protective in maintaining or improving BMD in elderly people. The addition of vitamin D and calcium supplementation to DHEA translated to a 4% improvement in BMD in women, comparable to the effect of bisphosphonate treatment. This effect was not observed in men.

DHEA replacement trials have not shown significant adverse events (e.g., increases in prostate-specific antigen), but much larger trials would be needed to establish safety and efficacy.

Aging is associated with a significant decline in the GH area under the curve (AUC), as well as the number and amplitude of nighttime GH peaks. These changes in GH secretion are associated with a steady decline in IGF-I after age 30 years. By age 65 years, most individuals have an IGF-I concentration that is near or below the lower limit of normal for healthy young individuals. The observed decline in the GH/IGF-I axis appears to occur above the level of the pituitary because long-term treatment with GH-releasing hormone (GHRH) and/or other GH secretagogues (GHS) mitigates much of the decline. The cause of the fall-off in axis activity is not clear but could be explained by age-related changes in GHRH, somatostatin, or ghrelin tone. Ghrelin appears to be the natural ligand for the GHS receptor. Although a close physiologic relationship exists between GH secretion and slow-wave sleep, it is unclear whether the altered GH/IGF-I axis is the consequence or cause of profound aging-related changes in sleep architecture.

Many of the body composition changes that occur with aging seem consistent with a GH/IGF-I–deficient state. Indeed, GH-deficient adults have many of the same physiologic abnormalities as older individuals, including the following:

 Reduced lean body and muscle mass

 Reduced strength and aerobic capacity

 Excess total, central, and intraabdominal fat

 High incidence of metabolic syndrome

 Reduced bone mass and density

 Reduced or absent slow-wave sleep

 High incidence of mood disturbance (depression)

Although GH therapy in younger GH-deficient patients improves body composition, bone density, and cholesterol levels and may decrease death, the efficacy and safety of therapy for otherwise “healthy” elderly persons are controversial. A systematic review of clinical trials of GH in healthy elderly persons concluded that therapy does increase IGF-I concentrations, although women may require higher doses of GH for longer periods than men to achieve physiologic replacement levels. Despite higher doses per kilogram of body weight, women do not consistently demonstrate the increase in lean body mass or decrease in fat mass that occurs in men. Further, translation into clinically significant changes in strength, function, bone density, and improved metabolic parameters has been difficult to demonstrate in either sex. GH treatment is associated with several important adverse events, such as a significant additional incidence compared with placebo of soft tissue edema (42%), arthralgias (16%), carpal tunnel syndrome (15%), gynecomastia (6%), new impaired fasting glucose or impaired glucose tolerance (4%), and new-onset of diabetes (4%).

The scant clinical experience of GH treatment for healthy elderly persons suggests that, although GH may minimally improve body composition, it does not improve other clinically relevant outcomes such as strength or function, and it is associated with high rates of adverse events. Furthermore, invertebrate and rodent models suggest that lower GH axis activity may be protective for longevity. On the basis of available evidence, GH cannot be recommended for use among healthy elderly persons. Large randomized controlled trials would be needed to determine the safety and efficacy of GH combined with an exercise intervention, combined with other sex hormones, effects in nonhealthy frail populations, and other replacement strategies such as GHRH.

Elderly persons often experience lack of sleep and feeling tired during the day. The cause may be the almost total loss of slow-wave sleep (stages 3 and 4). The periods of slow-wave sleep in younger individuals coincide exactly with the nighttime peaks of GH secretion. Indeed, animal and some human data suggest that appropriately timed GHRH supplementation may restart pulsatile GH secretion and stimulate slow-wave sleep. Limited evidence also suggests that long-term GHRH supplementation may improve cognitive function, specifically psychomotor and perceptual processing speed, as well as fluid memory.

As is the case for most hormones, distinguishing between the independent effects of age-related and body composition–related changes in the HPA axis is challenging. For example, morning cortisol levels tend to be lower and stress-induced HPA axis responsiveness tends to be greater in older as compared with younger adults, but such findings are also associated with central obesity (commonly found with aging; see earlier). However, several characteristics appear to be unique to aging. First, there is evidence for a phase advance characterized by an earlier morning cortisol peak. Second, the evening cortisol nadir appears to be higher in older persons, with a resulting compression of the diurnal amplitude. Third, glucocorticoid-mediated negative feedback is decreased. In total, mean 24-hour serum cortisol concentrations are 20% to 50% higher in both older women and men, likely reflecting the sum of alterations in glucocorticoid clearance, HPA axis responsiveness to stress, and central glucocorticoid-mediated negative feedback.

Whether an increase in the exposure to systemic and/or local/tissue (via 11-beta-hydroxysteroid dehydrogenase-1) glucocorticoids in elderly persons contributes to such age-related changes as central obesity, insulin resistance, decreased lean body mass, increased risk of fractures, decreased sleep quality, and poor memory (all common symptoms of cortisol excess) is an area of ongoing investigation.

Interpreting data from thyroid function studies in elderly subjects is difficult because evaluation is often complicated by increased prevalence of chronic disease and medication use. Nevertheless, serum thyroid-stimulating hormone (TSH) concentrations and distributions appear to increase with age, independent of the presence of antithyroid antibodies. This has been hypothesized to result from a decline in thyroid function or a recalibration of baseline TSH.

Serum reverse triiodothyronine (rT₃) concentrations appear to increase with age and the presence of disease. The presence of higher rT₃ was associated with a lower physical function status. Although serum free thyroxine (fT₄) levels tend to remain stable, older individuals with increased fT₄ levels were observed to have a lower physical function status and an increase in overall 4-year mortality. Total T₃ levels were inversely related with physical performance and lean body mass. These trends in thyroid hormone levels may indicate that it is beneficial to have lower activity of the thyroid hormone axis in old age.

Recognition of age-specific reference ranges would have important implications for defining subclinical hypothyroidism in elderly persons and treatment targets for thyroid hormone replacement.

Thyroid nodules increase with age, with an estimated prevalence of 37% to 57%. The risk of malignancy in a nodule also increases with age. The rate of carcinoma in a follicular nodule is increased in adults more than 60 years old and is higher in men than in women.

The most frequent cause of hyperthyroidism in older adults is toxic multinodular goiter, rather than Graves’ disease. Presenting symptoms of hyperthyroidism may be more atypical, with apathetic symptoms more common compared with younger patients.

Hypothyroidism increases significantly with age as a result of multiple conditions, including autoimmune thyroid dysfunction, use of medications, and nonthyroidal illness, which can lead to low serum thyroid hormone concentrations. The incidence of myxedema coma is also higher in older adults.

Subclinical hypothyroidism increases with age, but the actual incidence depends on the definition of upper limits of normal for TSH. For instance, in the United States, 15% of disease-free people who are more than 80 years old have TSH levels higher than 4.5 mIU/L, but if the definition were to change to TSH greater than 2.5 mIU/L, the incidence would be as high as 40%.

Subclinical hyperthyroidism is a predictor of mortality in elderly patients. Although subclinical hypothyroidism in individuals less than 65 years of age is associated with increased ischemic heart disease and cardiovascular mortality, this risk was not found in older adults. Further guidance on treatment of elderly persons with subclinical hypothyroidism should be generated by randomized controlled trials.

This leaves the question of what TSH concentration should be targeted in patients with frank hypothyroidism, given that epidemiologic studies have not revealed clinical concerns in patients with a TSH lower than 10 mIU/mL. The goal should be to avoid subclinical and frank hyperthyroidism, and a TSH concentration up to 10 mIU/mL is a reasonable target in elderly patients.

Among U.S. residents 65 years old and older, 10.9 million, or 26.9%, were known to have diabetes in 2010. When managing diabetes in elderly patients, treatment decisions should take into account the duration of diabetes and existent comorbidities such as heart disease or renal insufficiency, as well as polypharmacy and cost. A glycemic target of less than 8.0% may be more prudent in managing diabetes in patients with long duration of diabetes, established cardiovascular disease, limited life expectancy, and particular susceptibility to severe hypoglycemia. Large multicenter studies that aimed for intense glycemic control of hemoglobin A_1C (Hb A_1C) of less than 6.0% to 6.5% in older individuals showed no significant reduction in their primary combined cardiovascular end points but reported significantly more episodes of hypoglycemia in those patients with intensive blood glucose management. Thus, the risks of intensive control likely outweigh the benefits in an elderly population as a whole. Given the increasing complexity of glucose management in type 2 diabetes as new medications and drug classes are developed, a patient-centered treatment plan is necessary in reconciling glycemic management and optimal patient outcomes.

Special care is required in prescribing and monitoring drug therapy for older patients with diabetes. Metformin has traditionally been considered contraindicated at a creatinine value of 1.5 mg/dL or higher in men and 1.4 mg/dL or higher in women or in either gender when the estimated GFR (eGFR) is less than 60 mL/minute. It is important to assess the eGFR, which also takes age and body weight into consideration, because serum creatinine alone is often not an adequate reflection of GFR in older patients. Because of the age-related decline in renal function, the use of metformin is often discouraged in patients who are more than 80 years of age. However, more recent publications have noted that lactic acidosis is rare in patients with an eGFR of at least 30 mL/minute. Therefore, cautious use of metformin with close monitoring of renal function is a reasonable option in patients with an eGFR of 45 to 60 mL/minute, and lower dose metformin may be considered, again with close monitoring, in patients with an eGFR of 30 to 45 mL/minute. Pioglitazone should not be used in patients with congestive heart failure (New York Heart Association class III and IV); there is also an increased risk of peripheral fractures in elderly postmenopausal women using this medication. Furthermore, a possible increased risk of bladder cancer in patients taking pioglitazone is a concern. Insulin secretagogues, such as sulfonylureas, can cause hypoglycemia, and elderly patients may be particularly predisposed. Dipeptidyl peptidase-4 (DPP-4) inhibitors have an advantage of being able to be used in patients with renal impairment and are associated with less hypoglycemia. Insulin therapy requires good visual and motor skills and cognitive ability of the patient or a caregiver, and it can cause hypoglycemia. Hypoglycemia in older patients may be particularly difficult to identify and may be incorrectly diagnosed as irreversible cognitive impairment. Diabetes treatment can be improved in patients with visual impairments through the use of such devices as glucometers with large, easier-to-read screens, audio glucometers, magnifying glasses to help see syringes, or preloaded insulin pens.