Department of Obstetrics and Gynecology and Medicine, Columbia University College of Physicians and Surgeons, New York, New York 10032
Address correspondence and requests for reprints to: Michelle P. Warren, M.D., Columbia University, Department of Obstetrics/Gynecology, 622 West 168th Street, PH 1620, New York, New York 10032. E-mail:
Exercise induced amenorrhea occurs in athletes
with variable frequency, but the incidence varies from 5 to 25%,
depending on the level of competition and type of sport (Table 1). These women appear to have lost
normal pulsatility of LH and FSH due to environmental and metabolic
stresses. Thus the sensitivity of the reproductive system to
nutritional and other metabolic stresses is exemplified by the frequent
suppression of cyclicity seen in athletes. This condition is
interesting because of its neuroendocrine dysfunction and its
reversibility (1, 2).
|
The etiology of this disorder is the result of dysfunction at the
hypothalamic level. Specifically the gonadotropin releasing hormone
(GnRH) pulse generator appears to be affected. There appears to be
complete suppression of the normal pulsatile secretion of GnRH, which
generally occurs every 6090 min, as reflected in low levels of LH and
to a certain extent FSH. In some cases, there is a wide spectrum of
abnormalities from decreased frequency and amplitude of pulses to a
pattern of sporadic pulses or nocturnal entrainment of LH pulses (Fig. 1) (3). This may be due to influences of
weight loss or to dietary restriction on GnRH, or possibly both.
Accumulating evidence suggests that a critical metabolic or
"energy" balance is intimately tied to regulation of GnRH
pulsatility, and neuroendocrine adaptation to marginal energy intake is
the root of this problem in athletes. Studies attempting to reproduce
these abnormalities in normal women found that energy deprivation
affected LH pulsatility while exercise without energy deprivation did
not (4).
|
The great majority of these athletes will also exhibit some type of weight loss and dieting behavior. Usually they are not significantly underweight by modern standards, but they are generally thin and focused on their food and low fat intake. They are also often vegetarians or eat little red meat and may have iron deficient diets. The suppression of cycles may be mild, intermittent, and marked by fairly normal estrogen levels and positive withdrawal to progestins. Those who have a more exaggerated form of this condition, with low estrogen levels and in some cases very low gonadotropins (particularly LH), will generally be very thin and obsessed with their diets and athletic training. This group will not withdraw to estrogen.
Diagnosis: Spectrum of Clinical Presentations
Primary amenorrhea. The amenorrhea associated with exercising
is often difficult to diagnose, as it remains a diagnosis of exclusion.
Girls who exercise heavily at a young age may present with primary
amenorrhea, and all of the attendant differential diagnoses should be
considered. The very young athletes, particularly ballet dancers, who
start training as young as age 8 or 9, may have primary amenorrhea
until their early 20s. This problem is definitely related to the
exercise load (Fig. 2). However, this
group generally has normal growth and may continue to grow slowly until
that time. Generally, these individuals do not suffer from short
stature and will not have a concomitant delay in adrenarche as is seen
in constitutional delay of puberty. In fact, girls with delayed puberty
and menarche are known to be taller and have longer limb length (5).
The diagnosis of an exercise induced problem is always one of
exclusion. One athletic patient referred with "exercise" induced
amenorrhea had short stature and clitoromegaly and in fact had
congenital adrenal hyperplasia. Another patient, referred for the same
problem had weight loss associated with an obstructive ulcer. Although
central nervous system tumors are rare in this age group, particularly
the prolactin secreting tumors, they should also be considered.
|
Exercise induced amenorrhea presents generally in a young patient who prior to the onset of amenorrhea has had multiple metabolic and physiologic events that have inhibited the normal pulsality of LH and FSH secretion. These events are most likely subtle and do not manifest themselves until these insults are repeated and chronic and eventually menses stop. Subtle buffers that protect the reproductive system are affected: the most obvious is adequate weight or, as suggested by recent work, adequate body fat and leptin levels. The patients with exercise or weight loss related amenorrhea invariably are below or near so-called "ideal weight" and generally have low body fat and body mass index. They may have lost significant weight while exercising and may have obsessive eating problems including low fat and avoidance of red meat and all desserts. The diagnosis can be pinpointed in the competitive athlete with low gonadotropins, particularly low LH, normal prolactin, a negative pregnancy test, and no signs of androgenization such as acne, hair growth, or history of menarchial onset of irregular menses associated with signs of androgen excess. The typical patient with this problem will resist changing her behavior, particularly gaining weight or decreasing her exercise load. This is particularly true of the ballet dancer or competitive athlete. However, a change in eating habits, exercise patterns, and an increase in weight is the most effective method of reversing the problem.
Inadequate luteal phases. The athlete with exercise associated amenorrhea may initially have subtle forms of reproductive dysfunction manifested in shortening of the menstrual cycle. This has been reproduced in normal young women with intensive athletic training (6). Occasionally, the follicular phase may be prolonged. Thus these problems may present with irregular and/or frequent bleeding. After pregnancy is ruled out, this problem is handled by decreasing training or by use of oral contraceptives.
Health Risks: The Female Athlete Triad
Eating disorders and amenorrhea. The health of the athlete is at risk from three perspectives. Amenorrhea has already been discussed. The chronic dieting and amenorrhea may be a forerunner of an incipient eating disorder. Anorexia nervosa is common in athletes where low weight is desirable, and the increasing recognition of this problem has been the subject of much literature, both in the scientific and lay press (7). Bulimia or a typical eating disorder including obsessive avoidance of all food of high calorie value is also a problem.
Skeletal problems. The third, and probably most deleterious, risk is the impact on the skeleton, in particular osteoporosis and its secondary clinical effects, stress fractures and regular fractures. Although loss of bone in chronically hypoestrogenic athletes with secondary amenorrhea is a well recognized problem, a significant issue is the failure to reach peak bone mass, particularly in athletes with delayed menarche (8, 9). This phenomenon is probably a combination of genetics, intensive exercise, and negative energy balance. These adolescents will enter adulthood with what may be an irreversible bone mass deficiency and possibly increased fractures throughout their lives. Unfortunately, neither high calcium intake nor physical activity compensate for the lack of bone mass accretion in late maturing adolescents. Recent work suggests that this deficiency may be more pronounced in the spine than in the hips. However, stress fractures that are related also to a delay in menarche occur predominantly in weight bearing cortical bone (tibia, metatarsal, fibula, femur) and not trabecular bone. Risk of fracture is probably associated with bone density, but other factors may be operative (bone quality, total mass, training intensity, technique, and biomechanical factors).
Another problem related to the skeleton is the failure to mineralize bone with mechanical stress. This is seen in the feet of ballet dancers who have delayed menarche or secondary amenorrhea (10). Thus the physiochemical signal that stimulates bone to mineralize appears to be lacking in the aging female adolescents in the absence of sex steroids. The mechanism is unclear. On the otherhand, mineralization of the foot of the dancer with normal menses shows the expected increase in cortical density, producing striking x-rays of the foot (10).
A risk that has received little publicity and is also related to the skeletal system is the development of scoliosis in young athletes with delayed menarche, particularly ballet dancers, where the incidence is 2440% (11, 12). In contrast, there is a 3.9% incidence in white girls. The etiology of this condition is not well understood and may be related to the fact that the growth plates of the spine are the last to close. The recognition that eating disorders and osteoporosis occur in the setting of amenorrhea in the athlete has lead to the recognition of the so called "female athlete triad," which encompasses amenorrhea, osteoporosis, and eating disorders (13).
Some athletes present only with stress fractures. They may have recurrent episodes of secondary amenorrhea and recurrent bouts of weight loss and weight gain, or they may not have reproductive dysfunction. Athletes with menstrual irregularities however, are found not only to have a higher incidence of stress fractures but also multiple ones. Bone mineral density will be low but may occasionally be normal, and stress fractures may continue despite normal bone density, suggesting that bone quality is affected. These problems are particularly hard to treat, as hormone replacement has been suggested to prevent from further bone loss, but data suggests that these doses may not be effective. Calcium supplementation is always helpful. Ongoing studies suggest these individuals may have low bone turnover, but in-depth studies are lacking. It is imperative to examine the eating patterns in these individuals as they are often abnormal, and it is important to remember that stress fractures are also seen in patients without reproductive problems.
Diagnostic Problems
Patients who are suffering from underlying endocrinopathies affecting menstrual cyclicity are subject to problems with intense exercise, and the diagnoses of the causal factor may be difficult. As an example, women with polycystic ovarian disease may present with exercise induced amenorrhea if they have experienced weight loss below ideal weight with repetitive dieting and food restriction. In this case, LH levels may be low, but as the exercise induced problem resolves the underlying endocrinology becomes obvious, and acne and hirsutism with anovulation become manifest. LH levels may rise and testosterone levels that may have fallen during the exercise induced amenorrhea may rise to supranormal levels.
Subjects with exercise induced amenorrhea may experience a brief interlude of LH hyperactivity during recovery. If tested at this time, their LH levels may be high. Reasons for this appearance of high LH are unclear but may be similar to the increase in LH pulses sometimes seen with recovery from weight loss induced amenorrhea (14). Hyperreactivity to GnRH testing has unequivocally been demonstrated in this group. With recovery, LH pulses that have decreased in exercise induced amenorrhea may return to normal (3). Return of estradiol secretion is often associated with these events, and the patient may withdraw to progesterone or experience anovulatory bleeding or oligovulation before return of full menstrual cyclicity. This process may take up to a year. Rarely, the reproductive system does not recover, and patients remain amenorrheic.
A puzzling group are the women who return to an ideal weight and decrease their exercise load but continue to experience amenorrhea. Generally these women may still demonstrate some of the metabolic abnormalities seen when underweight: slow pulse, low blood pressure, sensitivity to cold, and rarely, hair loss. Undoubtedly, abnormal metabolic signals to the GnRH pulse generator maintains the amenorrheic status quo. These individuals have endocrine profiles similar to the athletes with low weight athletic amenorrhea and have suppression of menses despite cessation of athletic activity and normal weight. These women have been found to have abnormal eating patterns, similar to the athletes who develop stress fractures. Recent data published on the normal weight nonathletic hypothalamic amenorrhea suggest that the reproductive suppression may involve low leptin levels (15, 16, 17). Leptin, a hormone secreted by the fat cells and implicated in the initiation of puberty, may be lower in individuals who exercise and have low body fat. In addition, leptin is disproportionately lowered by fasting and is independently a regulator of metabolic rate. Athletes have low metabolic rates, particularly if they are amenorrheic. Thus, the altered metabolic state associated with marginal energy intake of the trained athlete can be associated with changes that signal the reproductive system (such as low leptin) and lead to menstrual dysfunction. Thus the athlete may represent the prototype of a model for which leptin acts as a metabolic signal to the reproductive axis. This is particularly intriguing as LH secretion is altered in this syndrome, and leptin replenishment has been shown to reverse the inhibitory effects of food restriction on primates. Leptin receptors have also been found in bone and may be important to osteoblastic function (18). Thus, the dieting in athletes, independent of the low estrogen secretion may independently affect bone metabolism and lead to osteoporosis by as of yet unidentified mechanisms.
Treatment
It is important to measure bone density, follow athletes with amenorrhea and/or fractures, and counsel them about their diets, fluctuations in weight, calcium intake, and in particular, the importance of keeping weight near normal levels and trying to lower exercise levels. Replacement of hormones is probably indicated, although hormone replacement therapy (HRT) in doses used for replacement in the menopausal years has not been effective in increasing bone mass. However, HRT may protect against further bone loss. Oral contraceptives have been helpful although more studies are needed and, in this young group, higher doses than used for the post-menopausal women may be needed. Young women during their reproductive years maintain higher levels of estrogen than that seen with HRT, suggesting that higher doses of hormones may be necessary. Additionally, the very competitive athlete, in particular the classical ballet dancer, may not be exposed to sunlight and may have a poor diet, so vitamin D should be added to calcium intake. If the bone loss involves a leptin pathway, other therapeutic modalities may be necessary. Until mechanisms are understood, preventive measures involve nutritional and behavioral changes. Certainly bulimic behavior and severe fasting should be avoided.
The first aim of therapy should be a decrease in exercise and an improvement in diet with weight gain. However, amenorrhea usually persists and may take over 6 months to reverse. Another problem of the athletic female is infertility. Amenorrhea usually responds to induction of ovulation particularly with exogenous gonadotropins, but more subtle problems may exist such as inadequate luteal phases. The athlete who wishes to become pregnant should cut back on her training regimen, and gaining weight may cure the inadequate luteal phases. The problem of inadequate luteal phases may be extremely subtle, and athletes attempting pregnancy should eliminate all strenuous training if experiencing infertility.
Inadequate luteal phases may also lead to frequent periods. If frequent bleeding is a problem, oral contraceptives can be used to maintain cyclicity and to suppress frequent bleeding.
A problem not frequently recognized is the menstrual dysfunction seen in athletes without low body weights (19). This is seen in such disciplines as swimming and is associated with increased LH/FSH ratios, higher androgen levels, and increased levels of DHEAS. Similar to the low weight athletic amenorrhea, this problem is easily reversed when athletic training is decreased. The endocrine pattern appears to be different, however, and is associated with a profile suggestive of anovulation and polycystic ovarian syndrome. This clinical syndrome has not been studied in depth, and it is not known if significant numbers of these individuals go on to develop polycystic ovarian syndrome. If polycystic ovaries do occur, this would be the first model suggesting that polycystic ovarian syndrome may, in certain instances, be induced by a stress such as competitive training. The estrogen levels appear to be normal in this group, and it is not known if osteopenia is a problem. Stress fractures have not been reported. Initial studies on rowers suggest that osteopenia is not a problem.
Overall, the health risks of the athletic female include psychological, reproductive, and skeletal disturbances. These problems appear to be reversible but are often difficult to detect without heightened awareness. The mechanisms are unclear, and more research is necessary.
Footnotes
1 Michelle P. Warren is Wyeth-Ayerst Professor of
Womens Health at Columbia University College of Physicians and
Surgeons and Medical Director, Center for Menopause, Hormonal
Disorders, and Womens Health, New York City, New York
10032.
Received February 5, 1999.
Revised March 24, 1999.
Accepted March 25, 1999.
References