Low Energy Availability in Female Athletes 

With the increase in the number of physically active women comes the need to understand the unique physiology of the female athlete. Although further research is needed to clarify issues such as the effect of exogenous ovarian hormones (oral contraceptives) on performance, many advances have been made.

One discovery has been that chronic energy deficit in the female athlete can cause musculoskeletal and reproductive dysfunction. Despite previous theories on the causes of menstrual dysfunction and the increased risk of stress fractures in the female athlete, studies have shown that the primary mechanism of menstrual disturbance in the female athlete is low energy availability. Even normally menstruating athletes can be in a state of low energy availability and can therefore experience deleterious effects on musculoskeletal health and performance.

Energy intake

The limiting factor for performance during training and competition in high-intensity sports of long duration is energy intake, especially carbohydrate intake, and a direct correlation exists between carbohydrate availability and reproductive and skeletal health. It might seem logical that increased energy expenditure, such as that during athletic training and performance, increases energy intake. However, for many athletes, particularly female athletes, this is simply not the case.

When this energy deficit is intentional, it is described as disordered eating and forms part of the female athlete triad, consisting of the following^{[1, 2, 3, 4, 5, 6, 7]} :

• Low energy availability
• Menstrual disturbance
• Low bone mineral density

Often, the deficit is unintentional, and physicians, athletic trainers, and even the athletes themselves can remain unaware of the condition and its potentially disastrous consequences.

Energy expenditure

The appetite of an athlete is not a reliable indicator of either energy balance or specific macronutrient requirements, because no biologic imperative to match intake to expenditure appears to exist.^[8] Hunger is actually suppressed for a brief period after a single episode of exercise at greater than 60% maximal oxygen consumption, or VO_2max.^[9]

Food deprivation increases hunger, but the same low energy availability, when caused by energy expenditure from exercise, does not increase appetite.^[10] In one study, a 20% increase in energy expenditure during 40 weeks of marathon training did not result in an increase in energy intake.^{[11, 12]} Even female monkeys with induced amenorrhea secondary to increased energy expenditure must be offered special treats in order to persuade them to consume enough food to restore normal menstrual function.^[13]

Other factors

Another factor in chronic energy deficiency in athletes is that body weight is not a reliable indicator of either energy or macronutrient balance. Because fat stores are associated with less body water than are protein and glycogen stores, the weight gain resulting from an increase in protein or glycogen stores more than counterbalances the weight loss resulting from the equivalent energy reduction in fat stores that occurs in low energy availability.

If disordered eating patterns are suspected, a therapist or psychiatrist familiar with treating eating disorders should be consulted immediately. For more information on the workup and treatment of eating disorders, please refer to the Medscape Reference article Female Athlete Triad.^{[1, 2]}

A good history should be obtained from all female athletes. In addition to the normal comprehensive history, certain components on which to focus include the athlete's nutritional, musculoskeletal, menstrual, endocrine/metabolic, psychosocial, performance, and medication history.

Nutritional history

Importantly, gather information about nutritional intake and eating patterns. The components of each athlete's diet are important in terms of the quantity of protein, carbohydrate, vitamins, and minerals consumed. Also important are the effect of training on an athlete's diet and the modification of the athlete's diet in times of increased training.

Musculoskeletal history

With the increase in risk of stress fractures in females with chronic energy deficit, a careful review of past and current musculoskeletal injuries in the female athlete should be conducted, with a focus on all stress fractures.Any injury that results in loss of training or competition time should be considered major.

Menstrual history

Other than fracture, the most likely manifestation of severe, chronic low energy availability in the female athlete is menstrual disturbance. Because many women do not volunteer information concerning menstrual disturbances, it should be specifically sought during all routine and sick visits by female athletes.

A complete menstrual history should be obtained. This includes the age at menarche, average length of menses, average time between menstrual periods, variations during times of increased training, and number of cycles per year. The possibility of pregnancy should be excluded in the case of amenorrhea. The patient's personal and family history of reproductive disorders, such as premature ovarian failure, should be discussed.

Endocrine/metabolic history

A history of and the risk for any endocrine abnormalities should be explored. Any personal and family history of thyroid disorders, pituitary disorders, and diabetes should be sought. Any family history of bone disease should be elicited.

Psychosocial history

Eating patterns should be discussed, and the Eating Disorder Inventory (EDI) may be used to screen for current and past disordered eating patterns. As with all patients, alcohol, tobacco, and drug use, as well as social support, depression, anxiety, and a history of abuse, should be determined.

Performance history

During sustained energy deficiency, an athlete's strength, power, and vertical jump all decline by approximately 20%.^[14] The athlete should be questioned as to whether she has noticed any change in strength or performance.

Medication history

All medications, dietary supplements, and herbal agents should be reviewed. Particular attention should be paid to medications, such as corticosteroids and anticonvulsants, that could affect bone health. The use of oral contraceptives should be elicited separately because patients often do not consider this a medication.

A typical comprehensive physical examination should be performed on all female athletes, with a focus on weight, percentage body fat, and thyroid function (for evidence of hypertrophy or irregularity).

For patients with menstrual irregularities, the physical examination should screen for pathologies that could cause metabolic and hormonal abnormalities. Particular attention should be paid to the parotid glands (for evidence of hypertrophy, as in bulimia), visual-field testing for pituitary adenoma, muscle strength, and the epidermis.

Skin findings might include evidence of hirsutism, vitiligo, or increased pigmentation of the palmar creases in adrenal insufficiency; easy bruising or stria in Cushing syndrome; warm, moist skin in hyperthyroidism; or any lanugo in anorexia.

A pelvic examination should be performed when appropriate and particularly when a patient presents with delayed menarche.

Urinary ketones

Because body weight alone is not a good indicator of body composition and energy availability, a noninvasive biomarker that reflects these variables should be identified. Until a better biomarker of body composition and energy availability is established, the measurement of urinary ketones might be the best indicator of sustained carbohydrate deficiency, because ketones are not present in urine when carbohydrates are available.

A baseline urinary acetoacetate measurement should be conducted by a trainer or team physician. Monthly measurements of urinary acetoacetate in times of increased training in normally menstruating athletes could be performed by the athlete herself or by the athletic trainer. The goal should be a complete absence of urinary ketones before and after a meal as well as before and after training.

In amenorrheic athletes or in those who are positive for ketones, daily keto-stick measurements can be made before and after meals, as well as before and after training, and energy intake can be increased until no evidence of acetoacetate is present in the urine and normal reproductive function is observed.^[15]

Other tests

Other helpful tests in patients with amenorrhea include the following^[16] :

• Beta human chorionic gonadotropin testing to rule out pregnancy
• A complete metabolic panel to assess electrolyte levels, renal function, and hepatic function and a complete blood cell (CBC) count to evaluate for anemia
• A thyroid panel to determine if the patient has hyperthyroidism or hypothyroidism

Testosterone, LH, FSH, estradiol, and prolactin tests are second-line options if a patient's reproductive function is not restored with a trial of increased energy intake or if the findings on physical examination and the patient's history suggest other causes of amenorrhea. If the patient has signs of hyperandrogenism (hirsutism), testosterone tests can help to assess for androgen excess.^[17]

An FSH level of approximately 40 mIU/mL indicates ovarian insufficiency. If a repeat value in 1 month confirms this finding and if the patient has experienced at least 4 months of amenorrhea, premature ovarian failure is confirmed. If the FSH level is 20-40 mIU/mL in a patient with disordered menses, the diagnosis is overt ovarian insufficiency, also known as prodromal premature ovarian failure.

LH levels are elevated in cases of 17,20-lyase deficiency; 17-hydroxylase deficiency; and premature ovarian failure.

When performing estradiol testing, draw and measure FSH values concomitantly. Serum estradiol levels within the reference range can be found intermittently, despite the presence of well-documented ovarian insufficiency. Finding a concomitantly elevated FSH level clarifies the issue. Serum estradiol levels fluctuate during the normal menstrual cycle. During the early follicular phase, levels may be less than 50 pg/mL. During the preovulatory estradiol surge, levels of approximately 400 pg/mL are not uncommon. The progesterone withdrawal test is not as valuable as the direct measurement of estradiol and FSH in determining ovarian health.

Prolactin levels in excess of 200 ng/mL are not observed except in the case of prolactin-secreting pituitary adenoma (prolactinoma). In general, the serum prolactin level is correlated with the size of the tumor. Psychotropic drugs, hypothyroidism, stress, and meals can also raise prolactin levels. Repeatedly elevated prolactin levels require further evaluation if the cause is not readily apparent.

If evidence from the patient's history or physical examination suggests the presence of a stress fracture, plain radiography should be performed, followed by 3-phase bone scanning if the radiographs are negative. Dual radiograph absorptiometry should be performed in athletes with multiple stress fractures.

Pelvic ultrasonography can be useful for determining the etiology of primary amenorrhea (eg, presence of ovaries, uterus). If abnormal pituitary function is suspected, thin-section magnetic resonance imaging (MRI) of the head through the sella turcica should be performed. Electrocardiography may show bradycardia, which is common in athletes, and a resting heart rate of less than 50 beats per minute should be explored with a baseline electrocardiogram.

When asked, 60% of women say that their menstrual cycle is 28 days long, but only approximately 12% of menstrual cycles are actually 28 days long.^[18] The length of the menstrual cycle varies from individual to individual, from cycle to cycle, from year to year, and from decade to decade. No one knows how much of the variation in the length and quality of menstrual cycles is due to environmental and behavioral factors, such as those that occur in athletic training, and how much is due to normal aging, inborn differences, disease, and random variation.

The median length of the menstrual cycle in the general population decreases from approximately 29 days in the first year after menarche to approximately 26 days after 40 years. If eumenorrhea is defined for college-age women to span 1 standard deviation around the mean, it ranges from 26-32 days, and 1 menstrual cycle in this range is most likely followed by another in the same range.^[19]

All types of menstrual disorders that occur in the general population also occur in athletes. However, they are more common in athletes, even young athletes, than in nonathletes. Approximately 31% of athletes who are not using oral contraceptives report symptomatic menstrual irregularities.^[20] If anything, exercise reduces dysmenorrhea, which is a common menstrual symptom and not a menstrual irregularity. More common irregularities in female athletes include primary or secondary amenorrhea, a shortened luteal phase, and oligomenorrhea.

On average, menarche appears to occur at a later age in athletes. Generally, girls in the United States tend to get their first menstrual period between ages 12 and 13 years. For female athletes, however, studies have found an average age of 13.6 years in track and field athletes,^[21] 14.2 years in Olympic volleyball candidates,^[22] 14 years in elite figure skaters, 12.9 years in elite Alpine racers, 13.4 years in competitive swimmers,^[23] and 15.6 years in elite gymnasts.^[24]

Although many studies have established that menarche often occurs later in athletes than in nonathletes, such retrospective surveys are inherently biased.^[25] Because the long bones continue to grow until increased estrogen levels close off the growth plates shortly after menarche and because longer bones are conducive to athletic success,^[26] later-maturing girls may be more athletically successful and therefore may choose to continue to participate in athletics, whereas earlier-maturing girls may be less successful and may therefore be socialized away from athletics.^[26] Although athletic training may delay menarche and even though animal research suggests that this is possible, no studies have validated this theory.

Large-scale studies have yet to be performed using the same 3-month definition of amenorrhea to directly compare data from female college athletes with that from general college-age women. Smaller studies of athletes using the same 3-month definition of amenorrhea have shown a prevalence as high as 44% in dancers and 65% in long-distance runners.^[27]

The incidence of luteal suppression and anovulation is high in regularly menstruating recreational and competitive athletes. Approximately 78% of regularly menstruating female runners have luteal suppression or anovulation in at least 1 month out of 3.^[28]

The term amenorrhea denotes the absence of menstrual cycles. A female has primary amenorrhea if she has not yet menstruated by age 16 years, even though she has undergone other normal changes that occur during puberty. The rate of primary amenorrhea in the United States is less than 0.1%.

The persistent absence of menstrual cycles beginning sometime after menarche is called secondary amenorrhea. Because ovarian follicular development, ovulation, and luteal function are not occurring in amenorrheic women, they are infertile. While recovering, however, they ovulate before menstruating (ie, before they know that their fertility is restored). Therefore, they should not rely on their amenorrhea for birth control purposes.^{[29, 30, 31, 32]}

The prevalence of amenorrhea strongly depends on how amenorrhea is defined. When the definition requires more months without menstrual periods, lower prevalence rates are reported. Studies of the general population have specified 3 months, because menstrual cycles longer than 90 days are extremely rare, even in the first and last decades of reproductive life. In large epidemiologic studies of college-age women in which amenorrhea was defined as no menstrual cycles for 3 consecutive months, prevalence rates of 2-5% have been reported.^{[33, 34, 35]}

If no evidence suggests musculoskeletal injury (eg, stress fracture) in an amenorrheic or oligomenorrheic patient, discontinuing or decreasing physical activity is not necessary to restore menstrual function. Animal studies have shown that an increase in energy intake is sufficient to restore menstrual function.^[15] Future studies on refeeding in amenorrheic females are needed to confirm that this is the safest and most effective treatment.

Different menstrual disorders can be symptoms of the same medical condition, and the same menstrual disorder can be a symptom of various conditions. Since the discovery in the late 20th century of progressive skeletal demineralization in amenorrheic athletes, most research has focused on amenorrhea in athletes. For inferential reasons, amenorrheic athletes have been compared with athletes who have highly regular menstrual cycles and with equally regular sedentary women. This research has identified undernutrition as the primary cause—and indeed the only demonstrated cause—of amenorrhea and luteal suppression in athletes.

The term oligomenorrhea (from the Greek oligos meaning few) is usually used to refer to menstrual cycles longer than 36 days. By far, most cases of oligomenorrhea occur in the first decade after menarche and in the last decade before menopause. Because the length of their cycles is so irregular, oligomenorrheic women may have difficulty conceiving, and paradoxically, they may also be at increased risk of an unintended pregnancy because of the difficulty of predicting the day of their next ovulation.

If no evidence suggests musculoskeletal injury (eg, stress fracture) in an oligomenorrheic patient, discontinuing or decreasing physical activity is not necessary to restore menstrual function. Animal studies have shown that an increase in energy intake is sufficient to restore menstrual function.^[15] Future studies on refeeding in amenorrheic women are needed to confirm that this is the safest and most effective treatment.

Findings from 2 studies suggest that in many athletes, the mechanism for oligomenorrhea may differ from that of amenorrhea. A study measuring the diurnal pattern of testosterone and pituitary hormone in female endurance athletes with menstrual disorders showed that in amenorrhea, hypothalamic inhibition due to energy deficiency appeared to play a major role; however, hyperandrogenism (increased testosterone secretion) seemed to be the major cause of oligomenorrhea.^[36]

The 24-hour hormone profiles in amenorrheic athletes showed decreased LH pulsatility and a peak amplitude of prolactin, as well as increased baseline levels of growth hormone and cortisol. In oligomenorrheic athletes, higher diurnal testosterone secretion was observed and levels of LH, prolactin, growth hormone, and cortisol were similar to those of regularly menstruating subjects.

High testosterone levels and oligomenorrhea (and occasionally amenorrhea) are symptoms of polycystic ovary syndrome, which is thought to be the most common cause of infertility in the United States. Therefore, just as it is plausible that later-maturing girls may self-select into athletics because of the rewards they receive for developing longer bones, it is also plausible that women with eating disorders may self-select into sports in which low body weight offers a competitive advantage.

Similarly, women with polycystic ovary syndrome may self-select into sports in which high testosterone levels offer an advantage. This possibility warrants focused research, because the treatment for polycystic ovary syndrome completely differs from that for undernourishment.

The term luteal suppression refers to an entirely asymptomatic, subclinical menstrual disorder that is evident only by measuring ovarian steroid hormone concentrations in the blood or urine over a number of weeks. In luteal suppression, follicular development progresses more slowly than usual; therefore, ovulation occurs later in the cycle. The luteal phase that follows may be short, and/or progesterone concentrations may be low. Often, the overall length of the menstrual cycle is indistinguishable from that in eumenorrhea. The incidence of a short luteal phase is 30-45% during the first decade after menarche, after which it declines to approximately 5%.^{[18, 37]}

The term anovulation refers to an asymptomatic subclinical menstrual disorder in which follicular development is so impaired that ovulation does not occur. Estrogen and progesterone levels are low, but enough estrogen is produced to stimulate some proliferation of the uterine lining, and bleeding occurs when the lining is sloughed. The incidence of anovulation declines from approximately 55% to less than 5% during the first decade after menarche and increases to approximately 20% in the last decade before menopause.^[18]

The issue of whether a female athlete should take oral contraceptives is a personal one; the importance of a drug's contraceptive and regulatory effects should be balanced against possible performance effects. Athletes should discuss this issue with their health care provider.

Studies have not shown that oral contraceptives have a major impact on performance, but several studies on monophasic and triphasic low-dose oral contraceptives have found a decrease in maximal oxygen consumption (VO_2max) in females from all athletic backgrounds after more than 1 month's use of these drugs. Athletes should be educated regarding this effect.^{[38, 39]} Because these same studies have shown a reversal of this effect within 1 month of the cessation of oral contraceptives, women who choose to take an oral contraceptive should be counseled that the decrease in VO_2max is likely reversible.

Oral contraceptives are prescribed to female athletes for several purposes, including contraception, cycle regulation, and control of dysmenorrhea (menstrual cramps). The oral contraceptive comes in many different brands with different synthetic hormones, doses, and dosing regimens. Those currently prescribed are usually given at doses lower than those of first-generation pills; therefore, they may have fewer adverse effects.

Oral contraceptives usually contain an estrogen and a progestin. The estrogen is commonly ethinyl estradiol, at doses of 20-50 mcg per pill. The progestin is often norethindrone, norgestrel, or levonorgestrel, with typical doses of 0.05-2.5 mg per pill.

The regimens are typically monophasic or triphasic. Triphasic dosing regimens more closely simulate the natural menstrual cycle and contain lower per-cycle progestin levels than do those of their monophasic counterparts.^{[20, 38]} The estradiol levels in most oral contraceptives contain approximately 3-5 times the estrogen and 1-2 times the progestin levels present in the normal menstrual cycle.^[40]

Because oral contraceptives might mask symptoms of amenorrhea induced by low energy availability, ketone measurements are even more essential in athletes taking oral contraceptives than they are in others.

Female athletes can be chronically energy deficient.^[15] With the exception of cross-country skiers, female endurance athletes consume approximately 70% as much energy and carbohydrates (controlled for body weight) as male athletes.^[41] Biochemical markers in female athletes indicate a mobilization of fat stores, slowing of the metabolic rate, and a decline in glucose utilization, with more extreme abnormalities in amenorrheic athletes and less extreme abnormalities in regularly menstruating athletes.^[15]

Studies in humans and monkeys have shown that this chronic energy deficiency causes reproductive disturbances in many female athletes. In 1998, Loucks and colleagues demonstrated that the stress of exercise did not suppress LH pulse frequency, because the disruption of LH pulsatility in exercising women could be prevented with dietary supplementation. On the other hand, low energy availability, caused either by an increase in exercise energy expenditure or by dietary energy restriction alone, did disrupt LH pulsatility.^[42] This low-energy state also suppressed levels of triiodothyronine (T3), insulin, insulinlike growth factor–1, and leptin and raised levels of growth hormone and cortisol in a pattern similar to those seen in amenorrhea and luteal suppression with eumenorrhea.^[42]

Menstrual function can also be restored by increasing energy availability. In female monkeys, amenorrhea induced by increasing exercise energy expenditure, with no reduction in dietary intake, was restored simply by means of dietary supplementation, without any moderation of the animals' exercise regimen.^[13]

Similar effects of low energy availability on the reproductive system have been noted in men.^[43] One week of dietary supplementation reversed disruptions of metabolic and reproductive hormones in US Army Ranger trainees despite continued exposure to strenuous exercise, sleep deprivation, cold, heat, injuries, infections, and other stresses. Therefore, exercise and other stresses appear to have no disruptive effect on the reproductive system apart from the impact of energy cost on energy availability.

Curiously, LH pulsatility is less disrupted in women who exercise than in women whose energy availability is reduced by exactly the same amount because of dietary restriction. This is surprising, because no one had previously suggested that exercise might be protective against menstrual disorders. On closer examination, working muscle in energy-deprived women who exercised reduced its glucose utilization, so that substantially more carbohydrate was available to the brain. This finding strengthens the hypothesis that reproductive function in women specifically depends on brain glucose availability.^[15]

The dose-response relationship between energy availability and LH pulsatility has also been investigated. LH pulsatility was found to have been disrupted below a threshold of energy availability of approximately 30 kcal/kg of lean body mass (LBM) per day. The dose-dependent effects on LH pulsatility most closely resembled the metabolic substrates glucose and beta-hydroxybutyrate and the metabolic hormones cortisol and growth hormone. These findings support the reported hypothesis that "reproductive function reflects the availability of metabolic fuels, especially glucose, which may be signaled in part by activation of the adrenal axis."^[15]

Stress fractures are common among female athletes. In one survey of competitive collegiate cross-country runners, 44% had experienced at least 1 stress fracture and 21% had suffered multiple stress fractures (Stanford B-Fit Web site). The incidence of stress fractures is greater in amenorrheic athletes, and bone density has been shown to be negatively correlated with the number of missed menstrual cycles since menarche.^[44]

Bone is a dynamic tissue that is constantly being remodeled. This remodeling is performed by osteoclasts (which resorb old bone) and osteoblasts (which form new bone) under the control of polypeptides, steroid hormones, thyroid hormones, cytokines, and growth factors.^[45] The balance between resorption and formation is usually coupled so that, in adults, bone mass remains stable.

The principal role of estrogen, through its action on osteoblasts and its indirect effect on osteoclasts, is to prevent bone resorption.^[46] In the past, some have theorized that the decreased bone densities observed in females with anorexia nervosa and in amenorrheic athletes was due solely to chronic hypoestrogenism. However, estrogen replacement in these individuals does not fully reverse the decrease in bone density.^{[47, 48, 49, 50, 51, 52]} This finding prompted researchers to investigate chronic undernutrition as an estrogen-independent mechanism for decreased bone mineral density in these patients.

The effects of low energy availability on bone health are evident even in normally menstruating sedentary women. In a landmark study, Ihle and Loucks demonstrated for the first time that bone formation is impaired within 5 days of the onset of low energy availability. At levels of energy deficit milder than that in bone resorption and at extreme energy restriction (10 kcal/kg LBM/day), increased bone resorption becomes uncoupled from decreased bone formation.^[53]

The findings of the study are applicable to female athletes, because normally menstruating athletes have reported energy availabilities of approximately 30 kcal/kg LBM/day, and amenorrheic athletes have reported energy availabilities of approximately 16 kcal/kg LBM/day.^[54]

A diversity of metabolic hormones—including carboxyterminal propeptide of type I procollagen and osteocalcin—are disrupted by all levels of energy restriction.^[55] The dose-response relationships of insulin, T3, and insulinlike growth factor–1 closely resemble those of propeptide of type I procollagen and osteocalcin and could be involved in mediating the effect of low energy availability on bone. In addition, as with N -telopeptide, estradiol is unaffected until energy restriction becomes severe. Because the primary role of estrogen in the skeletal system is to suppress osteoclast activity, this relationship is an expected finding.^[53]

Peak bone mass is a significant predictor of risk for the development of osteoporosis.^[46] The decrease in bone formation and increase in resorption (uncoupling) that occurs in a severe, chronic energy-deficient state is dangerous at any age. Because approximately 50% of peak bone mass is achieved in adolescence and is completed in most women by the end of the second decade,^{[56, 57]} the consequence of suppressing bone formation in adolescence can be disastrous.

Because appetite is not a reliable means of determining the energy needs of the female athlete and because exercise appears to suppress appetite, nutritional counseling is important for normally menstruating athletes and for athletes with menstrual disturbances.

Athletes need to eat by discipline, not by appetite. Studies have shown that a threshold of 20-30 kcal/kg LBM/day is needed for reproductive function and bone health, and a diet comfortably greater than 30 kcal/kg LBM/day is recommended.

Counseling should focus on how a patient's diet can provide close to 45 kcal/kg LBM/day in order for the patient to maintain reproductive and skeletal health. A balanced, nutrient-rich sample diet of 45 kcal/kg LBM/day tailored to individual weight or a general sample diet for weights of 55, 60, 65, and 70 kg on either a poster in the training room or a handout would be helpful. In addition, an estimation of the energy expenditure during a typical daily training regimen and the replacement of this expenditure are important.

Tables of energy expenditure in various athletic activities can be used for this calculation. For instance, a track coach could estimate the approximate energy expenditure during one practice (miles run), determine how many power bars or containers of yogurt are needed to replace the expended energy, and encourage athletes to eat these snacks after practice. A poster of snacks that replace, for instance, the energy lost during 1 hour of intermittent running (eg, during soccer practice) or after a run of a certain number of miles (eg, during track practice), could be displayed in the training room.

Athletes should be counseled that, if they underconsume, they are actually slowing their metabolic rate and that an increase in energy intake helps to restore their metabolic rate, prevent deleterious effects on reproductive and skeletal health, and even improve performance. Therefore, if the athlete's weight increases while her metabolic rate is adjusting, at least part of it is likely due to an increase in LBM, which even further improves performance.

For athletes with menstrual disturbances, energy intake should be even more closely monitored to ensure that it approximates 45 kcal/kg LBM/day along with the replacement of energy used during exercise. While intake is increased over a period of 1-2 weeks, electrolytes and hematocrit values should be monitored at least once a week during the transition.

Calcium intake is a key determinant of peak bone mass in adolescent women.^[46] Supplementation of calcium from adolescence through young adult life, with a recommended daily allowance of 1200 mg/day, is advised. In women aged 25-50 years, 1000 mg/day is recommended.^[46]

Vitamin K is a cofactor necessary for the gamma-carboxylation of several bone matrix proteins, one of which is osteocalcin. In one study of elite amenorrheic athletes, vitamin K supplementation induced an increase in bone formation markers and an even greater decrease in bone resorption markers.^[58] Therefore, supplementation with a multivitamin containing vitamin K might optimize bone health in female athletes. A multivitamin also contains other vitamins and minerals that serve as cofactors in the modification of the bone matrix, such as vitamin C, manganese, copper, and zinc.

Athletes who perform weight-bearing exercise have higher bone mineral densities than those who do not, but the advantage in bone mineral density tends to be site specific, depending on where the load has been applied. Studies have shown conflicting results in terms of the magnitude of force required to function as an osteogenic stimulus. It is unclear whether the increased bone mineral density from weight-bearing exercise persists into adulthood after the cessation of exercise.

The first published follow-up study of a 7-month, high-impact exercise program in prepubertal children after 7 months of detraining showed a persistent skeletal effect at the femoral neck.^[59] Follow-up studies of childhood exercise intervention after detraining will clarify this issue.

The addition of 50 vertical jumps per day has been shown to increase femoral and trochanteric bone mineral density by 2-3% in premenopausal women after 5 months^[60] and might prove valuable in creating an osteogenic stimulus in athletes, such as swimmers, who do not regularly load their skeleton during training.^[61]

Consultations include the following:

• A nutritionist can make recommendations for a balanced, nutrient-rich diet
• An obstetrician/gynecologist may be helpful for patients whose amenorrhea is resistant to refeeding
• A psychiatrist or counselor may be consulted in cases involving disordered eating (intentional energy [caloric] restriction)
• Regular follow-up with an athletic trainer, physiatrist, obstetrician/gynecologist, or team nutritionist or school dietary expert is necessary