Women and Exercise - Physiology and Sports Medicine

II Developmental Phases

This page intentionally left blank

CHAPTER 7 The Prepubescent Female ODED BAR-OR, M.D. PHYSIOLOGIC RESPONSE TO Response to Cold Climate SHORT-TERM EXERCISE High-Risk Groups for Heat- or Cold- Submaximal Oxygen Uptake Related Disorders Maximal Aerobic Power Anaerobic Power and Muscle GROWTH, PUBERTAL CHANGES, AND ATHLETIC TRAINING Endurance Muscle Strength COEDUCATIONAL PARTICIPATION IN CONTACT AND COLLISION TRAINABILITY SPORTS THERMOREGULATORY CAPACITY Response to Hot Climate r ReLecent years have seen an increasing interest in the physiologic response children to exercise. Such interest reflects the greater participation and success of prepubescents and adolescents in elite sports, as well as the recognition that physical exercise is relevant to the health of the nonathletic child. Although prepubescent athletes of both sexes engage in elite sports, it is pri- marily the females who have become extremely successful at the national and international levels. Such success is particularly apparent in gymnastics, figure skating, and swimming, in which prepubescents have been performing at levels that, a decade or two ago, were not considered feasible even for adults. To achieve such excellence, many female athletes have to practice as much as 4 to 6 hours per day and at high intensity. Such involvement and dedication has educational, psychosocial, medical, gynecologic, orthopedic, and physio- logic consequences. These have become a focus of research for sports scientists of various disciplines. Exercise-related research is oriented also toward the young nonathlete, healthy or ill.Study of the healthy child has been of interest, for example, tokin- anthropometrists, who are interested in growth patterns and the interrelation- ships between morphologic and functional changes; to epidemiologists, who assess the possible relationships between habitual activity during childhood and the risk of chronic disease in later years; to motor behaviorists, who study motor learning and skill acquisition; and to physiologists, who seek answers to such maturation-related issues as strength development, energy expenditureof locomotion, trainability,and thermoregulation. The relevance of exercise to the ill child has also generated growinginterest. 129

130 Developmental Phases Pediatric cardiologists and respirologists, ysis of relationships between fitness and for example, are using exercise for the as- growth in girls. sessment of children with such diseases as congenital heart defects, bronchial asthma, PHYSIOLOGIC RESPONSE TO and cystic fibrosis; an exercise prescription SHORT-TERM EXERCISE is incorporated into the management of the child with diabetes mellitus, obesity, mus- Differences in the response to short-term cular dystrophy, cerebral palsy, and cystic exercise (less than a 15-minuteduration) of fibrosis; and detrimental effects of exercise prepubescent and older females are sum- are studied in such conditions as aortic ste- marized in Table 7-1. Table 7-2 is a sum- nosis, dysrhythmia, primary amenorrhea, mary of gender-related differences in the re- and epilepsy. sponse of prepubescents to short-term exercise. The following discussion will high- This chapter is meant to focus on the light those characteristics of the prepubes- physiologic responses to exercise of the cent girl that have a direct relevance to her healthy prepubescent girl. Emphasis will be physical performance. given to differences among prepubescents, adolescents, and young adults. Differences Submaximal Oxygen Uptake will also be pointed out between the re- Typically for young girls, oxygen uptake sponses to exercise of girlsand boys. When- ever relevant, the implications to health of (calculated per body mass unit) while run- such differences will be pointed out. ning or walkingat any given speed is higher than in adolescent or adult females.1,22-24 A It is assumed that the reader has some 5.5-year-old girl, for example, who runs at 10 basic knowledge of exercise physiology. Ad- km/h, consumes about 46 mL of oxygen per ditional information on pediatric exercise kilogram of body weight per minute, com- physiology can be found in monographs,1-6 pared with 37 m L - k g - 1 - m i n - 1 in a 16-year- edited books,7-10 and proceedings of the Pe- old adolescent. Calculated per body surface diatric Work Physiology Group.11-20 Further- area, however, such differences disappear.25 more, a journal (Pediatric Exercise Sciences, Human Kinetics,Champaign, IL) is available The implication of such a high metabolic which is fully dedicated to the effects of ex- cost is that, at any walking or running speed, ercise in children. A recent monograph21 in- cludes a comprehensive, longitudinalanal- Table 7-1. SOME PHYSIOLOGICRESPONSES TO ACUTE EXERCISE:COMPARISON BETWEEN PREPUBESCENT GIRLS AND OLDER FEMALES Physiologic Function Typical for Girls (Compared with Older Females) O2 cost of walking/running min-1 Higher O2 uptake max, L • m i n - 1 Lower O2 uptake max, mL • k g - 1 • Higher Heart rate submax Higher Stroke volume submax Lower Cardiac output submax Lower Minute ventilation submax Higher Ventilatory equivalent submax and max Higher Peak anerobic power, watt Lower Peak anaerobic power, watt • kg-1 Lower Mean anaerobic power, watt Lower Mean anaerobic power, watt • k g - 1 Lower

The Prepubescent Female 131 Table 7-2. GENDER-RELATED however, there is little change in the Vo2max COMPARISON OF THE RESPONSE OF of girls up to the age of 10to 11years. During PREPUBESCENTS TO ACUTE EXERCISE the second decade of life, Vo2max per kilo- Physiologic Function Girls' Response gram decreases with age, such that it is ap- (Compared proximately 4 to 6 mL-kg~'-min-1 lower at with Boys) age 17 to 18 than at age 10to n.1,2,29'30 It has O2 cost ofwalking/running Similar been suggested that the lower Vo2max per O2 uptake max, L • min~' Somewhat lower kilogram in the pubertal girl is due to the de- O2 uptake max per kg body weight Somewhat lower crease in blood hemoglobin concentration, O2 uptake max per kg lean mass Similar secondary to menstrual blood loss. This, Heart rate submax Higher Heart rate max Similar however, does not explain the drop in Stroke volume submax and max Lower Vo2max per kilogram of body weight even Minute ventilation submax Similar before menarche. One reason could be the Minute ventilation max Somewhat lower increasing adiposity of many girls who ap- Peak anaerobic power, watt Somewhat lower proach puberty.31-33 Decreasing aerobic Peak anaerobic power, watt • kg-1 Lower power could also result from an age-related Mean anaerobic power, watt Somewhat lower decrease in spontaneous habitual activity in Mean anaerobic power, watt • kg-1 Lower the second decade of life.33-37 Although gender-related differences in maximal aerobic power are apparent pri- a young girl operates at a higher percentage marily after age 12to 13years, boys seem to of her maximal aerobic power and will fa- hliaevr eagaesso.1m,2,e33w,38h- a40t hInigahesrtuVdoy2mcoaxmepvaerninagt ear- tigue earlier than an older girl or a woman. the This may be the main reason why young maximal aerobic power of 6- to 16-year-old girls cannot compete on a par with their girls and boys who were tested on the cycle older counterparts in middle- and long-dis- ergometer, such gender-related differences tance running. Such a difference is virtually were eliminated when Vo2max (L-min-1) nonexistent during cycling.26-28 This sug- was plotted against lean leg volume rather gests that the biochemical-to-mechanical than against age.39 A similar pattern was ap- energy transfer efficiency in muscles is not parent among 8- to 16-year-old girls and lower at a young age, but young girls have a boys when W170(i.e., the mechanical power more \"wasteful\" gait, which increases their at which they cycle when their heart rate is mechanical output and metabolic demands 170 beats per minute) was plotted against during the gait cycle. No data are available body cell mass.41 It should be realized, how- on the age-related differences in the meta- ever, that when Vo2max is divided by lean bolic cost of swimming. The success of leg volume or lean body mass, preadoles- young girls in elite swimming would suggest, cent boys still have higher values than pre- however, that a proficient young swimmeris adolescent girls.33 A more precise determi- not less economical in her style than her nation of body composition is needed to tell older counterpart. whether gender-related differences in max- imal aerobic power of prepubescents are fully explained by the mass of their exercis- Maximal Aerobic Power ing muscles. Throughout childhood and adolescence, maximal aerobic power, as reflected by max- imal oxygen uptake (Vo2max), increases Anaerobic Power and Muscle with age.The Vo2max of 5-year-oldpre- Endurance schoolers is 0.80 to 0.90 L-min-1 compared with 1.1to l.5L-min-1 and 1.6 to 2.2L-min-1 High-intensity muscle contractions that for 10- and 16-year-old girls, respectively.1,2 cannot be sustained for more than 20 to 30 seconds are dependent primarilyon anaer- Calculated per kilogram of body weight, obic energy pathways. Examples of \"anaer-

132 Developmental Phases obic\" activities are short and long sprints in tive deficiency of anaerobic power in the running, skating, and cycling, as well as prepubescent. Children of both sexes have short slalom in downhillskiing.Until recent lower maximal blood lactate concentration years, this component of fitness received lit- than do adolescents and adults. It has been tle attention, compared with maximal aero- reported for boys, but not for girls, that cre- bic power and muscle strength. This re- atine phosphate and glycogen concentra- flected the paucity of reliable and valid tions in the resting muscle and, in particular, laboratory tests for peak muscle power and the rate of glycolysis in the contracting mus- local muscle endurance. Such tests are cur- cle are low before puberty. (For more de- rently available, using cycle ergometers or tails, see reference 2.) Based on animal stud- isokinetic machines. These have been ies, a relationship has been suggested added to the Margaria step-running test,42 between muscle lactate production and cir- which assesses peak muscle power but not culating testosterone. Whether this applies muscle endurance. The following informa- also to humans—females or males—has yet tion has been obtained using the Margaria to be shown. It can be assumed, however, test and the Wingateanaerobic test.43 that a low glycolytic capacity in prepubes- cents of both sexes is the main cause of their The ability of prepubertal girls and boys low anaerobic performance. to perform anaerobic tasks is distinctly lower than that of adolescents and young Muscle Strength adults. This was first shown for 8- to 73-year- Muscle strength, defined as the maximal old sedentary Italians: even when divided by body weight, the peak muscle power of force that can be exerted by a muscle or a the 8- to 10-year-old girls was only about group of muscles, is similar in girls and boys 60% that of the 20-year-old women.42Similar during their first decade of life.33,47 Strength results have been shown for Nilo-Hamitic is growth-dependent.33,47-49However, it does and Bantu African,44 British,39American,45 not increase linearly with the growth in and Israeli2 populations. In the last, peak body mass or stature. In girls, the main in- muscle power and muscle endurance of the crease in strength occurs during, a few arms and the legs were both lower in the months following, or even just before the young girls, even when corrected for differ- \"growth spurt\" (i.e., the year duringwhich ences in body weight. body height velocity is at its peak). In con- trast, the increase of strength in boys The aforementionedpattern is in contrast reaches its peak about 1 year after the to maximal aerobic power which, when cal- growth spurt.50,51 This difference, coupled culated per kilogram body weight, is higher with the earlier growth spurt in girls (about in the prepubescent girl than in the adoles- a 2-year difference), may explain why the cent or adult female. greater muscle strength of boys is usually not evident before age 11 or 12years. The mechanism for the lowanaerobic per- formance of prepubescent girls is not TRAINABILITY known. In a recent study performed in my laboratory on adolescent girls and boys,46 Does a prepubescent girl respond totrain- lean muscle mass of the upper limb ex- ing in the same manner as an adolescent or plained much of the variance in arm peak an adult female? This question is of utmost power and muscle endurance of the boys but relevance to the theory and practice of not of the girls. Performance of both prepu- coaching, but should be of interest also to bescent girls and boys in the Margaria test, the pediatric physiotherapist and the phys- even when corrected for fat-free mass, is iatrist who wish to apply physiologic prin- lower than that of adolescents and adults.44 ciples to rehabilitation. It is quite likely, therefore, that qualitative characteristics of the muscles, and possibly their neural control, would explain the rela-

The Prepubescent Female 133 To obtain definitive answers about train- Few studies are available on thetrainabil- ability (i.e., the ability of body systems to ity of muscle strength at different ages. Niel- adapt to repeated exercise stimuli) of differ- sen and co-workers49 trained 249 Danish ent age groups, one must conduct a longitu- girls aged 7 to 19 years for 5 weeks. One sub- dinal training study on these groups. Such a group did isometric knee extension, another design must satisfy two conditions: (1) the the \"vertical jump,\" and the third practiced initial fitness level of all groups must be sim- acceleration in sprints. As in adults, there ilar and (2) the training dosage must be was specificity in the responses: each sub- equated among the groups. group improved most in the specific strength (but not sprinting) task at which it Unfortunately, neither condition can be had been training.While the authors did not adequately satisfied. First, one cannot as- report the pubertal stage of the subjects, the sume that a 6-year-old girl who, for example, younger girls (less than 13.5years) im- sprints 50 m in 11.0 seconds has the same proved more than the older ones. Likewise, sprinting ability as a 16-year-old adolescent 8-year-old German girls improved their iso- who runs at the same speed. A better ap- metric arm strength more than did adults proach might be to use a physiologic crite- when given a similar training stimulus.57 rion for equating the initial fitness level. One Whether trainabilityof strength is related to cannot be sure, however, that a maximalaer- the pubertal stage has yet to be shown.Most obic power of 40 m L - k g - 1 m i n - 1 in a 6-year- research on muscle trainability of children old girl denotes the same aerobic fitness as is limited to boys. There are indications, an identical value in a 20-year-old woman. It however, that trainability duringprepuberty is also fairly difficult to equate training dos- is similar in boys and girls.58This is indirect ages. Can one assume, for example, that evidence that training-induced strength weight trainingat 70% of their maximalvol- gains can be achieved without the effect of untary contraction represents the same androgens. physiologic strain in a girl and a woman? THERMOREGULATORY Because of such methodologic con- CAPACITY straints, conclusions about the trainability of young girls are not definitive. Some pat- Most research on the thermoregulatory terns, however, seem to emerge. According characteristics of the exercising child is to several reports, when prepubescent girls based on studies in boys. (For a review, see take part in aerobic training, they respond reference 59.) Data are available, however, with little or no increase in maximal oxygen to suggest that girls are at a disadvantage, uptake, even though their athletic ability compared with women, when exposed to ei- may improve.52-55 This is unlike the re- ther hot or cold climates. Very few data are sponse to aerobic training of women, who available to compare the responses to heat increase their maximal oxygen uptake and and cold of prepubertal girls and boys. improve their athletic performance. Only a Response to Hot Climate few studies have suggested that prepubes- cent females do improve their maximal ox- Table 7-3 is a summary of the morpho- ygen uptake in response to aerobic train- logic and physiologic characteristics of pre- ing.56 pubescent females, as related to their ther- moregulatory capability. As discussed A major reason for the improvement of earlier, the smaller the girl, the higher her running performance in the absence of in- Vo2 per kilogramof body weightat any given creased Vo2max is the training-inducedim- walking or running speed. Because 75% to provement in running economy, which is 80% or more of the chemical energy during manifested by a decrease in the O2 cost of running. During adolescence, both aerobic power and running economy may improve with training.

134 Developmental Phases Table 7-3. MORPHOLOGIC AND PHYSIOLOGICCHARACTERISTICS OF PREPUBESCENT GIRLS AS RELATEDTO THERMOREGULATION Characteristic or Function Typical for Girls Implication for Thermoregulation (Compared with Women) O2 cost of running/walking Higher High metabolic heat Surface-to-mass ratio Larger Greater heat exchange with environment Onset of sweating Later Greater reliance on convective heat loss Sweating rate Somewhat lower Lower evaporative capacity Blood flow, peripheral vs. central Higher 1. Higher heat convection 2. Lower venous return muscle contraction is converted into heat, low sweating capacity), it also decreases the the metabolic heat load of the prepubescent venous return and stroke volume.60 The re- girl is higher (by as much as 5% to 20%)than sulting decrease in maximal cardiac output that of the adolescent or the adult, at equiv- is another explanation for the low ability of alent walking or running tasks. Such a differ- prepubertal girls to exercise intensively in ence imposes a greater strain on the young, hot climates. It should be added that, at any small girl's thermodissipatory system. given exercise level, even when performed in neutral environments, cardiac output in Another size-related difference is the young girls is somewhat lower than that of larger skin surface-area-per-mass ratio in women.61 the smaller individual.The rate of heat ex- change between the body and the environ- In summary, these geometric and physio- ment depends on this surface area. There- logic characteristics suggest that a priori fore, when the environment is warmer than young girls would tolerate exercise in hot the skin, the smaller the girl, the greater the climates less effectively than adolescent or heat gain (through conduction, convection, adult females. It has indeed been shown and radiation) per unit body mass. Thisdif- that, during extreme climatic heat, prepu- ference in heat gain becomes particularly bescent girls had to terminate their pre- important in extreme climatic heat. scribed walking task earlier than did young women.60'62 In thermoneutral environments, Evaporation of sweat is the main avenue on the other hand, there is no evidence that for heat dissipation during exercise, espe- young age or small body size is detrimental cially in hot climates. When ambient tem- to thermoregulation.63 perature exceeds skin temperature, evapo- ration is the only available means of heat As recently shown,64 the sweat of prepu- dissipation. Compared with women, prepu- bescent girls has a lower concentration of bertal girls have a slow onset of sweating sodium and chloride, and a higher concen- and a somewhat lower sweating rate while tration of potassium, than the sweat of ado- exercising in the heat,60which limit their ca- lescent females or of women. One possible pacity for evaporative cooling. This differ- implication of this difference is that the op- ence between prepubescents and adults timal electrolyte concentration of sports seems to be even more apparent among beverages may be different for prepubes- males. cent girls and for more mature females. Girls were found to respond to exercise in Response to Cold Climate the heat with a marked shift of blood from In most land-based sports, the rate of met- the central to the cutaneous vascular bed. Although greater skin blood flow facilitates abolic heat production exceeds heat loss, greater convection of heat from the body even when the environmental temperature core to the periphery (which, under certain is low. Such is the case, for example, in skat- climatic conditions, may compensate for a

The Pre:pubescent Female 135 ing and cross-country skiing. In other winter mented for college-age women,although the sports, such as downhill skiing or curling, findings for prepubertal girlswere inconclu- the rate of heat production may not be high, sive.62 but clothing usually prevents excessive heat loss. Hypothermia occurs not infrequently Hypohydration may often lead to heat-re- in such sports as mountain climbing, lated disorders. While data are not available snowshoeing, and even long-distance run- regarding the effects of hypohydration on ning at low intensity. There is, however, no the thermoregulation and health of prepu- epidemiologic evidence that prepubescent bescent girls, data on boys suggest that, for girls are more prone to hypothermia in a given level of hypohydration, children these events than are older females. have a greater rise in core temperature than do young adults.™ Conditions in which ex- In contrast, small individualsare at a dis- ertion may induce heat-related disorders tinct disadvantage during water-based ac- through hypohydration are diabetes melli- tivities. When swimming at a speed of 30 tus, diabetes insipidus, diarrhea, and vomit- m/rnin in 20.3°C water, 8-year-old girls (club ing. Prepubescent boys70 and girls (unpub- swimmers) had a drop in core temperature lished data from my laboratory), like adults, of as much as 2.5to 3.0°C and had to be taken undergo \"voluntary hypohydration\" when out of the water within 18 to 20 minutes they exercise for long periods (e.g., 1 to 2 owing to marked thermal discomfort. Their hours), even when fluids are available to 16- to 19-year-old clubmates swam for some them ad libitum. One group of young girls 30 minutes, with hardly any drop in core who are prone to hypohydration is those temperature and with little or no thermal who compete in judo and \"make weight\" discomfort.65 The reason for the cold intol- prior to competition. In some states where erance of the younger girls was their large elementary school girls compete in wres- surface area per mass, which facilitated con- tling, the same practice is probably fol- ductive heat loss (water having a heat con- lowed. ductivity at least 25 times that of air). The authors also found that the leaner girls had Lack of acclimatization to exercise in the a greater heat loss than those who had a heat is perhaps the most important factor thicker insulative subcutaneous fat layer. that predisposes an individual to heat-re- lated disorders. Data suggest that 8- to 10- High-Risk Groups for Heat- or year-old boys take longer than adults to ac- Cold-Related Disorders climatize to the heat.71 Nosimilarstudies are available for girls, but it makes good sense to Some girls are at a potentially high risk for ensure that young female athletes are well such heat-related disorders as heat exhaus- acclimatized to the heat before they are ex- tion or heat stroke, while others may be pected to train hard and perform well in prone to hypothermia. warm or humid climates. Evidence is available that girls with an- As for hypothermia, a small, lean girl who orexia nervosa have a deficient thermoreg- is immersed in water is at a greater risk than ulatory capability, both in the heat and in a larger girl or one with thicker subcutane- the cold.66,67 Patients with cystic fibrosis are ous adipose tissue. prone to heat-related disorders,68 possibly because of their abnormal sweating pattern. GROWTH, PUBERTAL Undernourished childrenare prone to both CHANGES, AND ATHLETIC hypothermia and hyperthermia.69 Obese in- TRAINING dividuals perform well and feel comfortable in cold climates but are less tolerant to ex- Trained prepubescent and adolescent ercise in the heat than their leaner counter- girls often have different morphologic and parts. Such intolerance has been docu- maturational characteristics from those of

136 DevelopmentalPhases their untrained counterparts. A question short legs, and higher body adiposity—drop often asked by coaches, physicians, and par- out because of unfavorablechanges in body ents is whether training per se affects mechanics. growth, development, and maturation. To obtain a definitive answer, one would need Based on reports from the late 1970s and to launch a prospective study in which non- early 1980s, delayed menarche (defined as athletic prepubescent girls are randomly as- occurring after age 15 years) was particu- signed to training and control groups and larly common among divers, figure skaters, then followed until after puberty. Such a gymnasts, and volleyball players.73 Menar- project has yet to be launched. Data avail- che is particularly delayed in those athletes able at present are based on cross-sectional who are engaged in high-dosage training. comparisons between athletes and nonath- Delayed menarche in athletes seems also to letes and among athletes of various special- correspond to delayed skeletal maturity.73 ties, or on retrospective analyses. The few longitudinal studies lack proper controls.72 Several factors, singly or in combination, The conclusions derived from such studies have been suggested to link delayed men- therefore are tentative at best and cannot arche to physical training. Among these are prove causality between training and a low percentage of body fat,80,81 insufficient changes in growth, development, and mat- calorie intake in conjunction with \"energy uration. drain,\"72 onset of training prior to menar- che,76 large sibship,82and emotional stress of The following are general comments, training and competition. It has also been based on such studies. (For detailed re- suggested,83 but has yet to be confirmed, that views, see Malina,73 Malina and co-work- hormonal changes which are associated ers,74 and Wells.75) Various female athletes, with chronic exercise may be a cause for primarily gymnasts, figure skaters, and bal- delayed menarche. In one study,72 low let dancers, mature later and are shorter serum gonadotropins—LH, particularly— than the nonathletic female population. Oth- were found in premenarcheal ballet danc- ers, notably swimmers, have little or no ers. Other endocrinologic studies are based delay in maturation and are often taller than on postmenarcheal athletes (see reference nonathletes.72-74,76-78 These data might sug- 84 for details). gest that the above morphologic and matu- rational differences are caused by training. In a comprehensive review on menarche Such conclusions, however, ignore prese- in athletes, Malina73 presented a two-part lection and a possible bias in the drop-out hypothesis on the possible relationship be- pattern. It is likely, for example, that those tween physical activity and delayed menar- girls with delayed puberty and short stature che. First is the preselection by body char- become preferentially attracted to such acteristics, in which the girl with a linear sports as gymnastics and figure skating, physique, long legs, and narrow hips (who is while the taller ones are more attracted to often also a late maturer) is attracted to competitive swimming. A recent retrospec- sports and eventually is successful in them. tive study79 has shown that 8- to 14-year-old Second is the \"socialization\" process, in female gymnasts who were shorter than the which early-maturinggirls tend to interact nonathletic population had been shorter socially in a nonsport environment with the even prior to having joined the gymnastics appearance of pubertal changes. Con- program. Similarly, swimmers, who, as a versely, the late maturers are more likely to group, were taller than their nonathletic find sports participation socially gratifying. counterparts, had been taller before train- ing. It is also possible that, within a groupof Indeed, preselection and the bias in drop- gymnasts, those females who mature ping out from athletics may explain the late early—and thus attain broad hips, relatively menarche of athletes as found in cross-sec- tional and retrospective analyses. One can- not ignore, however, the accumulatingdata on a more direct, possibly cause-and-effect,

The Prepubescent Female 137 relationship between intense sports partic- chronologic age around puberty, differences ipation and secondary amenorrhea. in body size and strength of early and late maturers within a gender group far surpass Although primary amenorrhea is a \"nor- the intergender differences. Nor is there any mal\" and common occurrence among ath- evidence to suggest that prepubescent girls letes, one should not overlook the possibil- are less capable of learning sport skills, are ity that it might reflect gynecologic or other less agile, or have less stamina than boys.87,88 hormonal abnormalities. (For details of rec- While not addressing specifically the pre- ommended investigations and of therapeu- pubescent girl, a recent review on orthope- tic approach, see Shangold in reference 85, dic issues in the young female athlete89 as well as Chapter 8.) points out the emergence of \"overuse inju- ries\" during the teens. It rejects, however, COEDUCATIONAL the notion that girlsare more prone to injury PARTICIPATION IN CONTACT than boys. AMD COLLISION SPORTS Based on anthropometric and fitness-re- Should prepubescent girls compete with lated considerations alone, therefore, pre- boys in contact (e.g., wrestling, basketball, pubescent girls can compete successfully soccer) and collision (e.g., football, ice with boys in contact and collision sports, hockey) sports? This issue has become and with no undue risk to health. An early highly controversial, attracting media atten- maturing girl, in point of fact, may have an tion, because of its medical, educational, edge over boys who are average maturers. It and cultural implications. The following seems as though matching of prepubescent comments are not meant to address the psy- and circumpubescent opponents by body chologic, sociologic, or ethical aspects of size and maturation level has more rele- this controversy but only some of the phys- vance to health than the separation intogen- iologic and medical aspects. der groups. A major issue is the added risk to health SUMMARY that participants of either sex group may incur owing to mixed participation. The The physiologic responses to exercise in main potential cause for such added risk is a the prepubescent girl are ofa similarpattern marked difference in body mass, strength, to those of the more mature female. There or skill among the participants. At age 9 to 12 are, however, some age- or development-re- years, body mass of girls is similar to, or lated differences in these responses. The even slightlygreater than, that of boys. Body submaximal 02 cost during walking or run- height at that age range is quite similar in ning is higher in the young girl, which boys and girls, and the difference in the causes a lower \"metabolic reserve\" and strength of various muscle groups is only early fatigability in endurance events. Like- about 1 to 2 kg in favor of the boys.51 This is wise, anaerobic muscle power and local to be contrasted with the increasingly muscle endurance are markedly lower in greater muscle strength of males—particu- prepubescents, who are therefore unlikely larly in the upper body—after puberty.86 The to compete successfully with their older attainment of such motor skills as throwing, counterparts in events such as jumping and kicking, catching, jumping, hopping, and sprinting. Girls are less-effective thermore- skipping during the first years of life is sim- gulators when exercising in the heat and in ilar in boys and girls. Throughout the pre- the cold. This has implications both to their pubertal years, these and other motor skills performance and to their health. Girls with seem to develop and improve at a similar obesity and anorexia nervosa are at special pace in both sex groups.87 risk for heat-related illness. Although more It should be realized that, at any given

138 Developmental Phases research is needed, it appears that the train- and Exercise. Acta Paediatr Belg (Suppl ing-induced improvement in maximal aero- 28):1, 1974. bic power is low before puberty. 15. Borms J, and Hebbelinck M (eds): Pediatric Work Physiology. Karger, Basel, 1978. A causal relationship among training, 16. Ilmarinen J, and Valimaki I: Pediatric Work growth, and maturation has yet to be estab- Physiology X. Springer Verlag, Berlin, 1983. lished. It seems, however, that the delayed 17. Lavallee H, and Shephard RJ (eds): Frontiers menarche in athletes may be in part a result of Activity and Child Health. Pelican, Quebec, of intense training. 1977. 18. Rutenfranz J (ed): Pediatric Work Physiology While coeducational participation in con- XII. Human Kinetics,Champaign, IL, 1986. tact and collision sports may be objected to 19. Thoren C (ed): Pediatric Work Physiology. on psychologic and societal grounds, there Acta Paediatr Scand (Suppl 213): 1, 1971. are no physiologic or medical reasons to 20. Frenkel R, and Szmodis I (eds): Children and ban such activities before puberty. Exercise. Pediatric Work Physiology XV. Sig- net, Budapest, Hungary, 1991. REFERENCES 21. Simons J, Beunen GP, Renson R, et al (eds): Growth and Fitness of Flemish Girls: The Leu- 1. Astrand PO: Experimental Studies of Physical ven Growth Study. Human Kinetics,Cham- Working Capacity in Relation to Sex and Age. paign, IL, 1990. Munksgaard, Copenhagen, 1952. 22. MacDougall JD, Roche PD, Bar-Or O, et al: Maximal aerobic power of Canadian school 2. Bar-Or O: Pediatric Sports Medicine for the children: Prediction based on age-related Practitioner: From Physiologic Principles to cost of running. Int J Sports Med 4:194, 1983. Clinical Applications. Springer Verlag, New 23. Robinson S: Experimental studies of physical York, 1983. fitness in relation to age. Int Z Angew Physiol Einschl Arbeitphysiol 10:251, 1938. 3. Godfrey S: Exercise Testing in Children.Ap- 24. Skinner JS, Bar-Or O, Bergsteinová V, et al: plications in Health and Disease. WB Saun- Comparison of continuous and intermittent ders, Philadelphia, 1974. test for determining maximal oxygen uptake in children. Acta Paediatr Scand (Suppl 4. Shephard RJ: Physical Activity and Growth. 217):24, 1971. Year Book Medical Publishers, Chicago, 25. Rowland TW, and Green GM: Physiological 1982. responses to treadmill exercise in females: Adult-child differences. Med Sci Sports Exerc 5. Malina RM, and Bouchard C: Growth, Matu- 20:474, 1988. ration, and Physical Activity. Human Kinet- 26. Bal MER, Thompson EM, Mclntosh EH, et al: ics, Champaign, IL, 1991. Mechanical efficiency in cycling of girls six to fourteen years of age. J Appl Physiol 6:185, 6. Rowland TW:Exercise and Children'sHealth. 1953. Human Kinetics, Champaign, IL, 1990. 27. Girandola RN, Wiswell RA, Frisch F, et al: Metabolic differences during exercise in pre- 7. Smith NJ (ed): Sports Medicine for Young and post-pubescent girls (abstract). MedSci Athletes. American Academy of Pediatrics, Sports Exerc 13:110, 1981. Evanston, IL, 1983. 28. Wilmore JH, and Sigerset PO: Physical work capacity of young girls 7-13 years of age. J 8. Boileau RA(ed): Advances in Pediatric Sport Appl Physiol 22:923, 1967. Sciences, Vol 1. Human Kinetics, Champaign, 29. Chatterjee S, Banerjee PK, Chatterjee P, et al: IL, 1984. Aerobic capacity of young girls. Indian J Med Res 69:327, 1979. 9. Bar-Or O (ed): Advances in Pediatric Sport 30. Drinkwater BL, Horvath SM, and Wells CL: Sciences, Vol 3. Human Kinetics, Champaign, Aerobic power of females, ages 10 to 68. J IL, 1989. Gerontol 30:385, 1975. 31. Forbes GB, and Amirhakimi GH: Skinfold 10. Gisolfi CV,and Lamb DR (eds): Perspectives thickness and body fat in children. Hum Biol in Exercise Science and Sports Medicine, Vol 42:401, 1970. 2. Benchmark Press, Indianapolis, 1989. 32. Karlberg P, and Taranger J: The somatic de- velopment of children in a Swedish urban 11. Bar-Or O (ed): Pediatric Work Physiology. community. Acta Paediatr Scand (Suppl):258, Wingate Institute, Natanya, Israel, 1973. 1977. 33. Sunnegardh J: Physical activity in relation to 12. Berg K, and Eriksson BO (eds): Children and Exercise IX. University Park Press, Balti- more, 1980. 13. Binkhorst RA, Kemper HCG, and Saris WHM: Children and Exercise XI. Human Kinetics, Champaign, IL, 1985. 14. Borms J, and Hebbelinck M (eds): Children

The Prepubescent Female 139 energy intake, body fat, physical work capac- et al: Training of \"functional muscular ity and muscle strength in 8- and 13-year-old strength\" in girls 7-19 years old. In Ber K and children in Sweden. Doctoral dissertation, Eriksson B (eds): Pediatric Work Physiology University of Uppsala, Uppsala, 1986. IX. University Park Press, Baltimore, 1980, p 34. Huenemann RL, Shapiro LR, Hampton MC, et 69. al: Teenagers' activities and attitudes toward 50. Beunen G, Malina RM, Van'Thof MA, et al: activity. J Am Diet Assoc 51:433, 1967. Timing of adolescent changes in motor per- 35. Ilmarinen J, and Rutenfranz J: Longitudinal formance. Symposium on Maturation and studies of the changes in habitual physical Growth, ACSM, Nashville, 1985. activity of schoolchildren and working ado- 51. Malina RM: Growth, strength and physical lescents. In Berg K, and Eriksson BO (eds): performance. In Stull GA (ed): Encyclopedia Children and Exercise IX, University Park of Physical Education, Fitness and Sports: Press, Baltimore, 1980, p 149. Training, Environment, Nutrition and Fit- 36. Telama R: Secondary School Pupils' Physical ness. Brighton Publishing, Salt Lake City, UT, Activity and Leisure-Time Sports, Vol III (in 1980, p 443. Finnish). Institute of Educational Research, 52. Bar-Or O, and Zwiren LD: Physiological ef- University of Jyvaskyla, Report No. 107, fects of increased frequency of physical edu- Jyvaskyla, Finland, 1971. cation classes and of endurance conditioning 37. Verschuur R, and Kemper HCG: The pattern on 9- to 10-year-old girls and boys. In Bar-Or of daily physical activity. In Kemper HCG O (ed): Pediatric Work Physiology. Wingate (ed): Growth, Health and Fitness of Teenag- Institute, Natanya, Israel, 1973, p 183. ers (Medicine and Sport Science, Vol 20). 53. Gilliam TB, and Freedson PS: Effects of a 12- Karger, Basel, 1985, p 169. week school physical education program on 38. Cooper DM,Weiler-Ravell D, Whipp BJ, et al: peak Vo2, body composition and blood lipids Growth-related changes in oxygen uptake in 7 to 9 year old children. Int J Sports Med and heart rate during progressive exercise in 1:73, 1980. children. Pediatr Res 18:845, 1984. 54. Mocellin R, and Wasmund U: Investigations 39. Davies CTM, Barnes C, and Godfrey S: Body of the influence of a running-training pro- composition and maximal exercise perfor- gramme on the cardiovascular and motor mance in children. Hum Biol 44:195, 1972. performance capacity in 53 boys and girls of 40. Yoshizawa S, Ishizaki T, and Honda H: Phys- a second and third primary school class. In ical fitness of children aged 5 and 6 years. J Bar-Or O (ed): Pediatric Work Physiology. Hum Ergol (Tokyo) 6:41, 1977. Wingate Institute, Natanya, Israel, 1973, p 41. Burmeister W, Rutenfranz J, Stresny W, et al: 279. Body cell mass and physical performance ca- 55. Yoshida T, Ishiko T, and Muraoka I: Effect of pPahcyistiyol(WEi17n0s)cohflsAcrhboeoiltpchhyisldiorel n3.1:I6n1t, Z Angew endurance training on cardiorespiratory 1972. functions of 5-year-old children. Int J Sports 42. Margaria R, Aghemo P, and Rovelli E: Mea- Med 1:91, 1980. surement of muscular power (anaerobic) in 56. Brown CH, Harrower Jr, and Deeter MF: The man. J Appl Physiol 21:1662,1966. effects of cross-country running on pre-ado- 43. Bar-Or O: The Wingate Anaerobic Test. Char- lescent girls. Med Sci Sports Exerc 4:1, 1972. acteristics and applications (in French). 57. Rohmert W: Rechts-links-Vergleich bei iso- Symbioses 13:157, 1981. metrichem Armmuskeltraining mit verschie- 44. DiPrampero PE, and Cerretelli P: Maximal denem Trainingsreiz bei achtjaringen Kin- muscular power (aerobic and anaerobic) in dren. Int Z Angew Physiol Einschl African natives. Ergonomics 12:51, 1969. Arbeitphysiol 26:363,1968. 45. Kuroski TT: Anaerobic power of children 58. Siegel JA, Camaione DN, and Manfredi TG: from ages 9 through 15 years. M.Sc. Thesis, The effects of upper body resistance training Florida State University, 1977. on prepubescent children. Pediatr Exerc Sci, 46. Blimkie JR, Roache P, Hay JT, and Bar-Or O: 1:145, 1989. Anaerobic power of arms in teenage boys and 59. Bar-Or O: Climate and the exercising child— girls: Relationship to lean tissue. Eur J Appl a review. Int J Sports Med 1:53, 1980. Physiol 57:677,1988. 60. Drinkwater BL, Kuppart 1C, Denton JE, et al: 47. Asmussen E: Growth in muscular strength Response of prepubertal girls and college and power. In Rarick L (ed): Physical Activ- women to work in the heat. J Appl Physiol ity, Human Growth and Development. Aca- 43:1046, 1977. demic Press, New York, 1973, p 60. 61. Bar-Or O, Shephard RJ, and Allan CL:Cardiac 48. Clarke HH: Physical and motor tests in the output of 10- to 13-year-old boys and girls Medford Boys' Growth Study. Prentice-Hall, during submaximal exercise. J Appl Physiol Englewood Cliffs, NJ, 1971. 30:219, 1971. 49. Nielsen B, Nielsen K, Behrendt Hansen M, 62. Haymes EM, Buskirk ER, Hodgson JL, et al:

14O Developmental Phases Heat tolerance of exercising lean and heavy 78. MalinaRM, Spirduso WW, Tate C,et al: Age at prepubertal girls. J Appl Physiol 36:566,1974. menarche and selected menstrual character- 63. Davies CTM:Thermal responses to exercise istics in athletes at different competitive lev- in children. Ergonomics 24:55,1981. els and in different sports. Med Sci Sports 64. Meyer F, Bar-Or O, MacDougall JD, and Hei- 10:218, 1978. genhauser JF: Sweat electrolyte loss during exercise in the heat: Effects of gender and 79. Peltenburg AL, Erich WBM, Bernink MJE, et level of maturity. Med Sci Sports Exerc al: Biological maturation, body composition, 24:776,1992. and growth of female gymnasts and control 65. Sloan REG, and Keatinge WR:Cooling ratesof groups of school girls and girl swimmers, young people swimming in cold water. J Appl aged 8 to 14 years: A cross-sectional survey Physiol 35:371,1973. of 1064 girls. Int J Sports Med 5:36,1984. 66. Davies CTM, Fohlin L, and Thoren C: Ther- moregulation in anorexia nervosa patients. In 80. Frisch RE, Wyshak G, and Vincent L: Delayed Borms J, and Hebbelinck M (eds): Pediatric menarche and amenorrhea in ballet dancers. Work Physiology. Karger, Basel, 1978, p 96. N Engl J Med 303:17, 1980. 67. Wakeling A, and Russel GFM: Disturbances in the regulation of body temperature in an- 81. Vanderbroucke JP, van Leer A, and Valken- orexia nervosa. Psychol Med 1:30, 1970. burg HA: Synergy between thinness and in- 68. Kessler WR,and Andersen DH:Heat prostra- tensive sports activity in delaying menarche. tion in fibrocystic disease of the pancreas Br Med J 284:1907,1982. and other conditions. Pediatrics 8:648, 1951. 69. Brooke OG: Thermal insulation in malnour- 82. Malina RM, Bouchard C, Shoup RF, et al: Age ished Jamaican children. Arch Dis Child at menarche, family size, and birth order in 48:901, 1973. athletes at the Montreal Olympic Games. 70. Bar-Or 0, Dotan R, Inbar O, et al: Voluntary Med Sci Sports 11:354,1979. hypohydration in 10- to 12-year-old boys. J Appl Physiol 48:104, 1980. 83. Brisson GR, Voile MA, DeCarufel D, et al: Ex- 71. Inbar O: Acclimatization to dry and hot envi- ercise-induced dissociation of blood prolac- ronment in young adults and children 8-10 tin response in young women according to years old. Ed.D. dissertation. Columbia Uni- their sports habits. Horm Metab Res 12:201, versity, New York, 1978. 1980. 72. Warren MP: The effects of exercise on puber- tal progression and reproductive function in 84. Shangold MM:Exercise and the adult female: girls. J Clin Endocrinol Metab 51:1150,1980. Hormonal and endocrine effects. In Terjung 73. Malina RM: Menarche in athletes: A synthe- RL (ed): Exercise and Sports Sciences Re- sis and hypothesis. Ann Hum Biol 10:1, 1983. views, Vol 12. Collamore Press, Lexington, 74. Malina RM, Meleski BW, and Shoup RF: An- MA, 1984, p 53. thropometric, body composition, and matu- rity characteristics of selected school-age 85. Shangold MM: Gynecological concerns in athletes. Pediatr Clin North Am 29:1305, young and adolescent physically active girls. 1982. Pediatrician 13:10, 1986. 75. Wells CL:Women, Sport and Performance— A Physiological Perspective. Human Kinet- 86. Montoye HJ, and Lamphier DE:Gripand arm ics, Champaign, IL, 1985. strength in males and females, age 10 to 69. 76. Frisch RE, Gotz-WelbergenAV,McArthur JW, Res Q Am Assoc Health Phys Ed 48:109,1977. et al: Delayed menarche and amenorrhea of college athletes in relation to age of onset of 87. Branta C, Haubenstricker J, and Seefeldt V: training. JAMA 246:1559, 1981. Age changes in motor skills during childhood 77. Malina RM, Harper AB, Avent HH, et al: Age and adolescence. In Terjung RL (ed): Exer- at menarche in athletes and non-athletes. cise and Sports Sciences Reviews, Vol 12. Col- Med Sci Sports 5:11, 1973. lamore Press, Lexington, MA, 1984, p 467. 88. Rarick GL, and Dobbins DA: Basic compo- nents in the motor performance of children six to nine years of age. Med Sci Sports 7:105, 1975. 89. Micheli LJ, and LaChabrier L: The young fe- male athlete. In Micheli LJ (ed): Pediatric and Adolescent Sport Medicine. Little, Brown, Boston, 1984, p 167.

8CHAPTER Growth, Performance, Activity, and Training during Adolescence ROBERT M. MALINA, Ph.D. THE ADOLESCENT GROWTH Significance of the Adolescent SPURT Plateau in Performance Body Size INFLUENCE OF TRAINING ON THE Body Composition TEMPO OF GROWTH AND MATURATION DURING MENARCHE ADOLESCENCE PHYSICAL PERFORMANCE AND Stature and Body Composition ACTIVITY DURING Sexual Maturation ADOLESCENCE Hormonal Responses Fatness and Menarche Strength Other Maturity Indicators Motor Performance Overtraining Maximal Aerobic Power Physical Activity Habits AAdolescenceis a period of transition from childhood to adulthood. It includes changes in the biologic,personal, and social domains that prepare the young girl for adulthood in her particular culture. Thus, the biologic changes that occur during puberty, or sexual maturation, do not occur in isolation; rather, they are related to other developmental events so that any consideration of this period of life must be done in a biosocial or biocultural context. Biologically, adolescence may be viewed as beginning with an acceleration in the rate of growth (i.e., an increase in size) prior to the attainment of sexual maturity, then merging into a decelerative phase, and eventually terminating with the cessation of growth. The latter is most often viewed as the attainment of adult stature. Sexual maturity and growth are thus closely related. The events that constitute this phase of the life cycle include changes in the nervous and endocrine systems that initiate and coordinate the sexual, physio- logic, and somatic changes; growth and maturation of the primary (ovaries, vagina, and uterus) and secondary (breasts and pubic hair) sex characteristics, leading to menarche and reproductive function; changes in size (i.e., the adoles- cent growth spurt); changes in proportions, physique, and body composition; and changes in the cardiorespiratory system, among others. The two most prom- 141

142 Developmental Phases inent outward features of adolescence (ex- starts in some girls as early as 7 or 8 years of cluding behavior) are accelerated growth age and in others as late as 12 or 13 years, and appearance of secondary sex character- while the age at maximum rate of growth in istics, which appear, on the average, during stature (PHV) occurs in some girls as early the second decade of life. However, the neu- as 9 or 10years of age and in others as late roendocrine and other physiologic events as 13 to 15 years.1 underlying growth and pubertal change Body Composition have been in progress for some time prior to the appearance of physical changes. The The fat-free mass (FFM) of girls, esti- time span accommodating the growth spurt mated from body density, increases from and puberty is thus wide. It can vary from 8 about 25 kg at 10 years to about 45 kg at 18 or 9 years through 17 or 18 years of age in years of age,whereas muscle mass, esti- girls, and in some cases may continue into mated from creatinine excretion, increases the early 20s. There is variation between in- from about 12 kg at 10 years to 23 kg at 18 dividuals in the time and rate at which the years.1 However, the major portion of structural and functional changes occur; change in FFMand muscle mass between 10 that is, the changes do not begin at the and 18 years occurs during the interval of same time and do not proceed at the same maximal growth (about 11 to 13 years in rate. girls). This interval includes PHV, whichoc- curs, on average, at about 12years of age in THE ADOLESCENT GROWTH girls. The adolescent gain in FFMand mus- SPURT cle mass during female adolescence is not, however, as intense as that in males, so that Body Size by late adolescence, females attain only From birth to adulthood, both height and about two thirds of the estimated mean val- ues reported for males. Peak velocities of weight follow a four-phased or double-sig- growth in arm and calf musculature occur, moid growth pattern: rapid gain in infancy on average, after PHV. and early childhood; slower, relatively con- stant gain in middle childhood; rapid gain Fatness also increases during adoles- during adolescence; and slow increase and cence, but estimates are highly variable. eventual cessation of growth at the attain- Densitometric estimates increase from 18% ment of adult size. Most dimensions of the body fat at about 10years of age to 23%at 18 body—sitting height, leg length, shoulder years.1 These estimates are adjusted for and hip breadths, limb circumferences, changes in the estimated chemical compo- muscle mass, and so on—follow a similar sition of FFM (i.e., density of FFM, potas- growth pattern. What varies is the timing, sium and water content of FFM)that occur tempo, and intensity of the adolescent during growth and are lower than those growth spurt in each. For example, maxi- based on adult chemical composition fig- mum growth (peak velocity) in leg length ures. At the time of the growth spurt, how- occurs early in the growth spurt, prior to ever, the rate of fat accumulation slows that for sitting height or trunk length, while down in girls. This is especially apparent on maximum growth in body weight occurs the extremities during the interval of PHV in after peak height velocity (PHV). girls. The timing of the growth spurt varies con- MENARCHE siderably among children. Most data are available for stature. According to data from The age at menarche is perhaps the most several longitudinal studies, the adolescent commonly reported developmental mile- growth spurt (i.e., the acceleration in rateof growth that marks the take-off of the spurt)

Growth, Performance, Activity, and Training during Adolescence 143 stone of female adolescence. It is, however, improves more slowly.1 This pattern is in a rather late maturational event. Menarche contrast to the marked acceleration of occurs after maximum growth in stature; the strength development during male adoles- average difference between menarche and cence, so that sex differences in muscular PHV in a number of studies is about 1.2to 1.3 strength are considerable. years.1 The relationship between strength devel- Menarche in Americangirls occurs, on av- opment and the growth spurt and sexual erage, just before the 13th birthday. How- maturation in girls is not as clear as in boys. ever, there is variabilitywithin the U.S.pop- Maximum strength development occurs, on ulation. In the National Health Examination the average, after peak height and weightve- Survey in the 1960s, the median ages at men- locity in boys, the relationship being better arche were 12.5years for black girls and 12.8 with weight than with height.1,5 In girls, the years for white girls.2 The median age at available longitudinaldata vary. In the Oak- menarche in American girls has not land (California) Growth Study, the time of changed, on average, since the 1950s.3 Esti- maximum strength development (a compos- mates for a number of European samples ite strength score of right and left grip and vary between 12.5 and 13.4 years.1,4 pushing and pulling tests) does not closely correspond to the growth spurt in stature, In contrast to population surveys of men- and a significant percentage of girls experi- arche, in which the average age for the pop- ence peak strength gains prior to PHV.6 Peak ulation is estimated mathematically on the strength gain precedes peak weight gain in basis of the number of girls in each age more than half of the girls, and follows peak group who have attained menarche, many weight gain in only about one fourth. On the studies of athletes and of the influence of other hand, in the study of Dutch girls training on the age at menarche use the ret- (Growth and Health of Teenagers), peak de- rospective method. This approach relies on velopment of strength (arm pull test) oc- the memory of the individual and thus has curs, on average, one-half year after PHV the limitationof error in recall. (the same time as it occurs in Dutch boys).7 The maximum gain in strength at this time is PHYSICAL PERFORMANCEAND about 6.0 kg/y in girls, which contrasts with ACTIVITY DURING a maximumgain of 12.0 kg/y in Dutch boys.7 ADOLESCENCE The data for Dutch girls are not expressed relative to peak gain in body weight. Characteristics of the adolescent growth spurt and sexual maturation, and of interre- Early-maturing girls are slightly stronger lationships among indices of sexual, skele- than late-maturinggirls of the same chron- tal, and somatic maturity, are reasonably ologic age during early adolescence, about well documented. Changes in physical per- 11 through 13 years.1 The differences be- formance and activity during female adoles- tween girls of contrasting maturity status, cence are less well documented. The data however, do not persist and are no longer are largely cross-sectional, with but fewlon- evident by 14to 15years of age. Further, the gitudinal observations spanning the imme- differences in muscular strength between diate prepubertal and pubertal years. girls of contrasting maturity status during adolescence are not as marked as those be- Strength tween early- and late-maturing boys. The strength advantage of girls advanced in ma- Muscular strength improves linearly with turity status between 11 and 13 years re- age from early childhood through about 15 flects the larger body size of early maturers, years of age in girls, with no clear evidence since strength is positively related to body of an adolescent spurt. After age 15, strength mass. When strength is expressed per unit body weight, early maturers have less

144 Developmental Phases strength per unit body weight than late-ma- of high- and low-performing girls indicated turing girls; this difference persists through that the superior performerswere about 0.5 adolescence.1 year less mature skeletally and 0.4 year later in menarche.12This trend is apparent in elite Motor Performance female athletes (i.e., skilled performers), Average performances of girls in a variety who tend to be later in age at menarche and delayed in skeletalmaturation.13,14 of motor tasks (dash, standing long jump, jump and reach, distance throw, and others) Maximal Aerobic Power improve more or less linearly from child- Absolute maximal oxygen uptake (mL/ hood through about 13 or 14 years of age, followed by a plateau in the ability to per- min) has a growth pattern in girls similar to form some tasks and a decline in others.1,8,9 that for motor performance:it increases lin- In most tasks, the average performancesof early with age from 7 years through 13to 14 girls fall within one standard deviation of years in untrained girls, and then declines the boys' averages in early adolescence. slightly.15 In contrast, in untrained boys, it After 13 to 14 years of age, however, the av- increases linearly with age through adoles- erage performances of girls are often out- cence, so that by 16 years of age the differ- side the limits defined by one standard de- ence between maximaloxygen uptake in un- viation below the boys' mean performance. trained boys and girls is about 56%.When Overhand throwing performance is an ex- expressed relative to body weight (mL- ception; few girls approximate the throwing kg-1 • min-1), aerobic power declines with performances of boys at all ages from late age from 6 through 16 years in untrained childhood on. girls, but is more or less constant in un- trained boys. The slope of the regression in Longitudinal data relating the motor per- girls declines from a value of 52.0 mL-kg-1 formance of girls to the timing of the adoles- min-1 at 6 years of age to 40.5 mL-kg-1 cent growth spurt are not available. Cross- min-1 at 16years. Values for untrained boys sectional analysis of longitudinaldata does at corresponding ages are 52.8 and 53.5 mL- not suggest adolescent spurts in the motor kg-1 • min-1, respectively, yielding a negligi- performances of girls. Performances in a va- ble sex difference of 1.5% at 6 years, but a riety of motor tasks show no tendency to considerable difference of 32% at 16 years.15 peak before, at, or after menarche (which occurs, on average, about 1year after PHV); The sex difference in aerobic power per rather, performances are generally stable unit of body weightat 16 years of age is prob- across time.5 Among boys, on the other ably related to sex differences in body com- hand, motor performances show rather position. The aerobic power of girls per unit clear adolescent spurts. Maximal gains in of body weight is approximately 77% of the functional strength and power tests (flexed value for boys. This percentage is not too dif- arm hang and vertical jump) occur, on av- ferent from estimates of lean body and mus- erage, after PHV, whereas maximal gains in cle mass in late adolescence; that is, girls at- speed tests (shuttle run, speed of hand tain, on the average, only about two thirdsof movement) and flexibility (sit and reach) the values for boys. The increase in relative occur before PHV.10 fatness associated with the sexual matura- tion of girls probably contributes to the sex Correlations between skeletal and sexual difference in aerobic power per unit of body maturity and motor performance in girls are weight. low and, for many tasks, negative. The latter suggests that later maturation is more often Absolute aerobic power (mL/min) shows associated with better motor performance a clear adolescent spurt in both girls and in girls, whereas the opposite is more often boys, which on average occurs close to that true in boys.1,11,12 For example, a comparison for stature.16 This reflects the growth of

Growth, Performance, Activity, and Training during Adolescence 145 heart and lung functions in proportion to through 9 (12 to 14years), and 11.8 hours in overall body size.1 Given the size differences grades 11 and 12 (15 to 17years). Although between early- and late-maturing girls, the the data suggest a trend, more specific former have a slightly larger absolute aero- changes with age cannot be examined. In a bic power, especially during early adoles- mixed-longitudinal sample of Dutch girls,20 cence. When expressed per unit of body the average number of hours per week spent weight, however, relative aerobic power is in physical activity with an average energy higher in late maturers.17 expenditure of 4 metabolic equivalents (METs) or more declined from 9.6 hours at Aerobic power responds positively to 12 to 13 years to 8.1 hours at 17 to 18 years. training, so that absolute and relative maxi- The earlier adolescent years were not con- mal oxygen uptakes are greater in trained sidered. than in untrained girls at all ages. The differ- ences between trained and untrained girls Intensity is a critical variable when con- are greatest duringadolescence. It is also in- sidering physical activity. In the mixed-lon- teresting to note that trained girls and boys gitudinal Dutch study, girls aged 12 to 13 differ by only 24% for absolute and 18% for participated, on the average, in only 4.0 h/ relative oxygen uptake at age 16, in contrast wk of activities of medium intensity (7 to 10 to comparable differences of 56%and 32%in METs), and 0.5/h/wk in activities of heavy untrained boys and girls of the same age.15 intensity (10+ METs).By 17to 18years, the corresponding hours per week were 1.5and Studies of aerobic power seldom control O.3.20 Clearly,the majority of the activities of for the maturity status of the subjects, and these girls were of light intensity. the fewstudies that do are largely limited to boys. Correlations between skeletal age and Given the type of data available, it is diffi- aerobic power are generally low,15 but the cult to make inferencesabout activity habits association between body mass and skeletal during the adolescent growth spurt and sex- maturity confounds the relationship.1 ual maturation, as well as about possible ef- fects of rapid growth and maturation on Physical Activity Habits activity habits. The figures do suggest, how- Physical activity is a major component of ever, that most adolescent girls are not get- ting sufficient regular physical activity to the daily energy expenditure. Energy expen- maintain a high level of aerobic fitness. diture in free-living childrenand youth is dif- ficultto measure, and the fewavailable stud- Significance of the Adolescent ies are limited to rather small samples with Plateau in Performance narrow age ranges, and largely to boys.18 Standardized questionnaires, interviews, Data relating the physical performance of and diaries are often used to estimate phys- girls to the timing of the growth spurt and ical activity habits in large samples of sexual maturation are not extensive. A ques- youngsters, usually 10 years of age and tion that merits more detailed study is the older. The data, however, are largely de- relative flatness of the performance curves scriptive and do not consider growth and of girls during adolescence. That is, their maturity characteristics. Results of several level of performance shows little improve- surveys of European, Canadian, and Ameri- ment in many tasks after 13 to 14 years of can youth indicate a slight decline in time age, and in some tasks it actually declines. Is spent in physical activity by girls during ad- this trend related primarily to biologic olescence.18 In the UnitedStates survey,19for changes in female adolescence (e.g., sexual example, the average weekly time engaged maturation, fat accumulation, physique in physical activity outside of school physi- changes), or is it related to cultural factors cal education was 11.5 hours in grades 5 and (e.g., changing social interests and expec- 6 (10 to 11 years), 12.5 hours in grades 7 tations, pressure from peers, lack of moti-

146 Developmental Phases vation, limited opportunities to participate of this maturational event may bepro- in performance-related physical activities)? grammed by conditions early in life and Most likely both biologic and cultural fac- not necessarily by those conditions that tors are reflected in the trend. Thus, the may be operating at or about the time of pu- overall age-related pattern of physical per- berty.24-25 formance during female adolescence may change with the recent emphasis on and op- A question of concern, therefore, is the portunity for athletic competition for young role of intensive training for sport and per- girls, and the wider acceptability of women haps of the stress of competition on the tim- in the role of athlete. ing and tempo of growth and sexual matu- ration during adolescence. It should be INFLUENCE OF TRAINING ON obvious thatphysical activity is only one of the THE TEMPO OF GROWTH AND many factors that may influence growth and MATURATION DURING maturation. ADOLESCENCE Stature and Body Composition Under adequate environmental condi- Regular physical training has no apparent tions, the timing of the adolescent growth spurt and sexual maturation is genetically effect on statural growth. It is, however, a determined. However, these processes can significant factor in the regulation of body be influenced by environmental factors. The weight and composition, specifically fat- delaying effects of chronic undernutrition ness. Changes in response to short- or long- are well documented. Socioeconomic varia- term training programs largely reflect fluc- tion in growth and maturation is evident in tuating levels of fatness, with minimal or no some societies but not in others.1 Criteria of change in FFM. The role of regular activity in socioeconomic status, of course, vary from the development of adipose tissue cellular- country to country, but data from industri- ity and subcutaneous fat distribution is not alized countries indicate inconsistent clearly established.26 trends in ages at PHV and menarche relative to indices of socioeconomic status. Another Regular training is a significant factor in factor related to age at menarche is the num- the growth and integrityof skeletal and mus- ber of children in the family. Girls from cle tissues. Changes in bone tissue include larger families tend to experience menarche greater mineralization, density, and mass. later than those from smaller families, and Training-associated changes in muscle tis- this applies to athletes as well as nonath- sue are generally specific to the type of pro- letes. The estimated effect of each additional gram followed. Strength or resistance train- sibling on the age at menarche ranges from ing is associated with hypertrophy, whereas 0.11 to 0.22 years in several samples of ath- endurance training is associated with in- letes and nonathletes.21 creases in oxidative enzymes. The direction of responses to training in growing individ- Stressful life events are also significant. uals is similar to those observed in adults, They are especially evident in the growth but the magnitude of the responses varies.26 and maturation of youngsters experiencing disturbed home environments,22 and in the The persistence of beneficial training ef- \"unusually 'fractured' curves of growth and fects on adipose and muscular tissues de- pubertal development in girls translated to pends upon continued activity. In contrast, unfamiliar boarding schools at various evidence is accumulating that excessive times in puberty.\"23 Studies of secular training associated with altered menstrual change in menarche suggest that the timing function (seebelow and Chapter 9) and diet contributes to bone loss in some athletes.27,28 Thus, there may be a threshold for some ad- olescent athletes: regular traininghas a ben- eficial effect on the integrity of skeletal tis-

Growth, Performance, Activity, and Training during Adolescence 147 sue up to a point, but excessive activity may between years of training before menarche alter menstrual function and have a negative and age at menarche, a moderate correla- influence on bone mass. tion that accounts for only about 28%of the sample variance. Correlation does not imply Sexual Maturation a cause-and-effect sequence, however; the Longitudinal data on the effects oftraining association is more likely an artifact. The older a girl is at menarche, the more likely on sexual maturation of girls (andboys too) she would have begun her training prior to are lacking, and the available cross-sec- menarche, and conversely, the younger a tional data do not indicate a significant effect girl is at menarche, the more likely she of training on sexual maturation. Much of would have begun training after menarche the discussion of training and sexual matu- or would have a shorter period of training ration is based on comparisons of later prior to menarche.29 It could also be that mean ages at menarche of athletes with later maturation is a factor in a girl's deci- those of the general population, with the in- sion to take up sport, rather than the train- ference that intensive training for sport\"de- ing causing the lateness.13 Further, athletes lays\" menarche.13 The menarcheal data are as a group tend to be rather select, and other generally consistent with observations of factors knownto influence menarche are not breast and pubic-hair development and considered in the analysis. skeletal maturity of young athletes engaged in figure skating, ballet, gymnastics, and It has also been suggested that menarche track—that is, they develop later.14 How- occurs later specifically in those disciplines ever, girls training for sport at prepubertal that emphasize low body weight, such as ages are not necessarily representative of ballet and gymnastics.31 Emphasis on low those who are successful at later ages, who body weight may involve dietary practices in turn constitute the samples of athletes that adversely influence maturation, so that upon whom most menarcheal data are it would be difficult to partition dietary from based. Also, Title IX legislation has influ- training effects. In addition, such sports enced sport opportunity for girls and tend to have rather rigorous selection cri- women, so that many now continue to train teria, which are often applied early in child- and compete through the college years. In hood and which favor the morphologic the not-too-distant past, on the other hand, characteristics of the late-maturing girl. Fi- many young girls stopped trainingand com- nally, data for elite university-level athletes peting at 16 or 17 years of age.The oppor- indicate later mean ages at menarche in ath- tunity provided by Title IX most likely has letes across several sports that differ con- influenced the composition of the female siderably in training load and emphasis on athlete population at the college level, par- body weight: diving, track and field, swim- ticularly in swimming. The age at menar- ming, tennis, golf, basketball, and volley- che in college-age swimmers in recent ball.21 estimates21-29 is considerably older than that of elite swimmers about 20 years ago,13 and Nevertheless, two questions merit consid- this is in contrast to the advanced pubertal eration. First, are regular, intensive, prepu- status and skeletal maturity often observed bertal training for sport and regular compe- in age groupswimmers.14 tition sufficiently stressful to prolong the prepubertal state and in turn delay the ado- Although not the first to suggest that train- lescent growth spurt and sexual maturation? ing may delay menarche, Frisch and Second, do intensive training for sport and colleagues30 concluded that for every year a the stress of competition during the adoles- girl trains before menarche, her menarche cent growth spurt and sexual maturation will be delayed by up to 5 months. This con- produce conditions that are sufficiently ad- clusion is based on a correlation of +0.53 verse to influence the progress and thus the timing of these maturational events?

148 Developmental Phases Hormonal Responses association with only \"mild\" growth stunt- The suggested mechanism for the associ- ing, for example, has been reported in pre- menarcheal ballet dancers.34 The dancers ation between training and later menarche were delayed in breast development, men- is hormonal. It is suggested that intensive arche, and skeletal maturation,which would training and perhaps the associated energy suggest a prolonged prepubertal state. How- drain influence circulating levels of gonad- ever, they were not delayed in pubic hair de- otropic and ovarian hormones, and in turn, velopment. menarche. Lower plasma levels of estrone, testoster- Exercise is an effective means of stressing one, and androstenedione have been ob- the hypothalamic-pituitary-ovarian axis, served in 11-year-old prepubertal gymnasts producing short-term increases in serum than in swimmers of the same age and ma- levels of all gonadotropic and sex steroid turity status, but plasma gonadotropin and hormones.32,33 Other factors also influence dehydroepiandrosterone-sulfate (DHEAS) hormonal levels, including diurnal varia- levels did not differ in the two samples. On tion, state of feeding or fasting, emotional the other hand, plasma levels of the seven states, and so on, and these need to be con- hormones assayed did not differ between sidered. Further,virtually all hormones are early pubertal (stage 2 of breast develop- episodically secreted, so that studies ofhor- ment) gymnasts and swimmers, although monal responses based on single serum the latter were an average of 0.5year older.35 samples may not reflect the overall pattern. Both the prepubertal and early pubertal What is needed are studies in which 24-hour gymnasts had been training regularly for a levels of hormones are monitored or in longer period than the swimmers. The two which actual pulses are sampled every 20 groups of gymnasts had been training since minutes or so in response to exercise. Oth- 4.8 and 5.0 years of age, respectively, erwise, the evidence from the available whereas the two groups of swimmers had studies on the hormonal response to exer- been trainingsince 7.2 and 8.0 years of age. cise is inconclusive. The similar levels of DHEAS in the prepu- bertal gymnasts and swimmers suggests a It should be noted that the majority of hor- similar stage of adrenarche, although the monal data do not deal with chronic gymnasts had been training for a signifi- changes associated with regular, intensive cantly longer period. This observation thus training. Further, the data are largely de- does not support the suggestion that train- rived from samples of postmenarcheal ing delays adrenarche and prolongs the pre- women, both athletes and nonathletes, who pubertal state.36 Moreover, recent evidence are physiologically quite different from the does not support the view that secretion of maturing girl. What is specifically relevant adrenal androgens triggers sexual matura- for the prepubertal or pubertal girl is the tion.37 Early childhood growth data for the possible cumulative effects of hormonal re- two groups of athletes suggest physique dif- sponses to regular training. The hormonal ferences. Since 3 years of age, the gymnasts responses are apparently essential to meet had been shorter and lighter than Dutchref- the stress that intensive activity imposes on erence data, whereas the swimmers had the body. Do they have an effect on the hy- been taller and heavier. Midparental heights pothalamic center, which apparently trig- (height of mother and height of father, di- gers and coordinates the changes that initi- vided by 2) and weights were also less in the ate sexual maturation and eventually gymnasts than in the swimmers, and the menarche? Such data are nowlacking. groups did not differ in socioeconomic sta- tus.35 Hormonal data for prepubertal or puber- tal girls involved in regular trainingare lim- Changes in basal levels of hormones in as- ited, and the results are variable and incon- sociation with training in young athletes clusive. Low gonadotropin secretion in

Growth, Performance, Activity, and Training during Adolescence 149 may be significant. Similar basal levels of occur.41 Accordingly, intensive, regular ACTH, cortisol, prolactin, and testosterone training functions to reduce and maintain have been reported during a 24-week train- fatness below the hypothesized minimal ing season in small samples of premenar- level, thereby delaying menarche. The crit- cheal and postmenarcheal competitive ical weight or fatness hypothesis has been swimmers 13 to 18years of age.38 During the discussed at length by many authors,21-42 and season, ACTH levels gradually increased, the evidence does not support the specific- prolactin levels tended to increase, and tes- ity of weight or fatness, or of a threshold tosterone levels decreased, whereas corti- level, as the critical variable for menarche to sol levels showed a variable pattern in the occur. combined sample. As expected, basal estra- diol levels differed between the premenar- Other Maturity Indicators cheal and postmenarcheal swimmers, but Since indicators of sexual maturity are both groups experienced a decrease in basal levels during the first 12 weeks of training, reasonably well related to indicators of skel- followed by a rise at 24 weeks. Basal levels etal and somatic maturity during adoles- of estradiol at the start of training and after cence,1 it seems logical to consider the 24 weeks of trainingdid not differ in the pre- effects of training on other maturity indica- menarcheal swimmers, whereas the basal tors. If the hormonal responses to regular level after 24 weeks was lower than at the training are viewed as important influences start of training in the postmenarcheal on sexual maturation, one might expect swimmers.38 them to influence the growth spurt, which occurs a year or so before menarche, and A role for B-endorphins in the amenor- skeletal maturation around the time of men- rhea of runners and, in turn, in later menar- arche. (For example, epiphyseal capping che in athletes has been postulated. Admin- and fusion are influenced by gonadal hor- istration of naloxone, an opiate receptor mones, among others.) antagonist, to amenorrheic athletes, for ex- ample, results in a marked increase in lu- Regular physical activity, including train- teinizing hormone (LH).39Responses of nor- ing for sport, has no apparent effect on other mal prepubertal girls and boys to naloxone indices of biologic maturation used in under basal conditions are different from growth studies. Age at PHV is not affected by those of adults, however.40 Naloxone appar- training, whileskeletal maturation is neither ently does not have an effect on LHsecretion accelerated nor delayed by regular training in children. A study of the effects of nalox- for sport during childhood and adoles- one during exercise conditions in children cence.1,2,26 might be enlightening, but ethical concerns make collection of such data difficult. Overtraining The issue of overtraining—thatis, exces- Fatness and Menarche A corollary of the suggestion that training sive training without adequate time for re- covery—must be considered, since a signif- delays menarche is that changes in weight icant number of adolescent girls (and boys) or body composition associated with inten- are involved in intensive training for sport. sive training may function to delay menar- Overtraining can be short-term or chronic, che; that is, training may delay maturation in and when it is chronic, it results in an array young girls by keeping them lean. This idea of behavioral, emotional, and physiologic is related to the criticalweight or critical fat- symptoms.43 Data for adults indicate weight ness hypothesis, which suggests that a cer- loss, decreased performance, and slow re- tain level of weight (about 48 kg) or fatness covery after training.Reduction in bothFFM (about 17%) is necessary for menarche to and fat mass probably accompany weight

150 Developmental Phases loss, and a reduction in efficiency and max- needed in which youngsters of both sexes imal working capacity accompany the de- are followed from prepubescence through crease in performance. Implications for puberty, in which several indicators of growing girls should be obvious. The behav- growth and maturity are observed, and in ioral, emotional, and physiologic complica- which both training and other factors known tions of overtraining have the potential to to influence growth and maturation are negatively influencegrowth and maturation. monitored. SUMMARY REFERENCES 1. Malina RM, and Bouchard C: Growth, Matu- Variation in the timing, tempo, and mag- nitude of the adolescent growth spurt is con- ration, and Physical Activity. HumanKinetics siderable. Although on the average girls Publishers, Champaign, IL, 1991. enter and complete the growth spurt earlier 2. MacMahon B: Age at menarche, United than boys, adolescent gains in FFM and States. Vital and Health Statistics, Series 11, muscle mass in girls are not as great as in No. 133, 1973. boys. Thus, young adult women attain about 3. Malina RM: Research on secular trends in two thirds of the estimated FFMand muscle auxology. Anthropol Anz 48:209,1990. mass levels of young adult men. In contrast, 4. Danker-Hopfe H: Menarcheal age in Europe. absolute and relative fatness increase more Yrbk Phys Anthropol 29:81,1986. in adolescent girls. 5. Beunen G, and Malina RM:Growth and phys- ical performance relative to the timing of the Menarche is a relatively late pubertal adolescent spurt. Exerc Sport Sci Rev 16:503, event that usually occurs a year or so after 1988. maximum growth in stature during the ado- 6. Faust MS: Somatic development of adoles- lescent spurt. In American girls, menarche cent girls. Mon Soc Res Child Dev42(l), 1977. occurs, on average, near the 13th birthday. 7. Kemper HCG,and Verschuur R: Motor per- formance fitness tests. In Kemper HCG (ed): Strength, motor performance, and abso- Growth, Health and Fitness of Teenagers. S lute aerobic power improve during adoles- Karger, Basel, 1985, p 107. cence, but the average performance levels 8. Branta C, Haubenstricker J, and Seefeldt V: tend to reach a plateau between 13 and 15 Age changes in motor skills during childhood years of age. Well-defined growth spurts in and adolescence. Exerc Sport Sci Rev 12:467, the strength and motor performances of ad- 1984. olescent girls are not clearly apparent. How- 9. Haubenstricker JL, and Seefeldt VD: Acquisi- ever, maximal aerobic power shows a defi- tion of motor skills during childhood. In See- nite spurt near the time of PHV. Trained girls feldt V (ed): Physical Activity and Well- have higher performance levels than do un- Being. American Alliancefor Health, Physical trained girls, and girls who are later in sex- Education, Recreation and Dance, Reston, ual and skeletal maturity tend to be better VA, 1986, p 41. performers. 10. Beunen GP, Malina RM, Van't Hof MA, et al: Adolescent Growth and Motor Performance: Under adequate environmental circum- A LongitudinalStudy of Belgian Boys. Human stances, the timing of the growth spurt and Kinetics Publishers, Champaign, IL, 1988. sexual maturation is genetically deter- 11. Beunen G, Ostyn M, Renson R, et al: Skeletal mined. The evidence that regular training maturation and physical fitness of girls aged before sexual maturity may delay matura- 12 through 16. Hermes (Leuven) 19:445,1976. tion of girls is not convincing. 12. Espenschade A: Motor performance in adolescence. Monogr Soc Res Child Dev The stress of training and competition as 5(1):1940. a factor that influences growth and biologic 13. Malina RM: Menarche in athletes: A synthe- maturation needs more systematic and con- sis and hypothesis. Ann Hum Biol 10:1, 1983. trolled study. Prospective studies are 14. Malina RM: Biological maturity status of young athletes. In Malina RM (ed): Young Athletes: Biological, Psychological, and Edu- cational Perspectives. Human Kinetics Pub- lishers, Champaign,IL, 1988, p 121.

Growth, Performance, Activity, and Training during Adolescence 151 15. Krahenbuhl GS, Skinner JS, and Kohrt WM: 28. Warren MP, Brooks-GunnJ, Hamilton LH, et Developmental aspects of maximal aerobic al: Scoliosis and fractures in young ballet power in children. Exerc Sport Sci Rev dancers. N Engl J Med 314:1348,1986. 13:503, 1985. 29. Stager JM, Robertshaw D, and Miescher E: 16. MirwaldRL, and Bailey DA:Maximal Aerobic Delayed menarche in swimmers in relation to Power. Sp>ort Dynamics, London, Ontario, age at onset of training and athletic perfor- 1986. mance. Med Sci Sports Exerc 16:550,1984. 17. Kemper HCG, Verschuur R, and Ritmeester 30. Frisch RE, Gotz-Welbergen AV,McArthur JW, JW: Maximal aerobic power in early and late et al: Delayed menarche and amenorrhea of maturing teenagers. In Rutenfranz J, Mocellin college athletes in relation to age of onset of R, and Klimt F (eds): Children and Exercise training. JAMA 246:1559, 1981. XII. Human Kinetics Publishers, Champaign, IL, 1986, p 213. 31. Warren MP, and Brooks-Gunn J: Delayed menarche in athletes: The role of low energy 18. Malina RM:Energyexpenditure and physical intake and eating disorders and their relation activity during childhood and youth. In De- to bone density. In Laron Z, and Rogol AD mirjian A (ed): Human Growth:A Multidisci- (eds): Hormones and Sport. Raven Press, plinary Review. Taylor and Francis, London, New York, 1989, p41. 1986, p 215. 32. Shangold MM: Exercise and the adult female: 19. Ross JG, Dotson CO,GilbertGG,et al: The Na- Hormonal and endocrine effects. Exerc Sport tional Children and Youth Fitness Survey: Sci Rev 12:53, 1984. After school physical education ... Physical activity outside of school physical education 33. Keizer HA, and Rogol AD: Physical exercise programs. J Phys Educ Rec Dance 56:77, and menstrual cycle alterations: What are the 1985. mechanisms: Sports Med 10:218, 1990. 20. Kemper HCG, Dekker HJP,Ootjers MG, et al: 34. Warren MP: The effects of exercise on puber- Growth and health of teenagers in the Neth- tal progression and reproductive function in erlands: Survey of multidisciplinary longitu- girls. J Clin Endocrinol Metab 51:1150,1980. dinal studies and comparison to recent re- sults of a Dutch study. Int J Sports Med 4:202, 35. Peltenburg AL, Erich WBM, Thijssen JJH, et 1983. al: Sex hormone profiles of premenarcheal athletes. Eur J Appl Physiol 52:385,1984. 21. Malina RM: Darwinian fitness, physical fit- ness and physical activity. In Mascie-Taylor 36. Brisson GR, Dulac S, Peronnet F, et al: The CGN, and Lasker GW (eds): Applications of onset of menarche: A late event in pubertal Biological Anthropology to Human Affairs. progression to be affected by physical train- Cambridge University Press, Cambridge, ing. Can J Appl Sport Sci 7:61,1982. 1991, p 143. 37. Wierman ME, and Crowley WR Jr: Neuroen- 22. Patton RG: Growthand psychological factors. docrine control of the onset of puberty. In In Mechanisms of Regulation of Growth, Re- Falkner F, and Tanner JM (eds): Human port of the 40th Ross Conference on Pediatric Growth, Vol 2. Plenum, New York, 1986, p 225. Research. Ross Laboratories, Columbus, OH, 1962, p 58. 38. Carli G, Martelli G, Viti A, et al: The effect of swimming training on hormone levels in 23. Tanner JM: Fetus into Man. Harvard Univer- girls. J Sports Med Phys Fit 23:45,1983. sity Press, Cambridge,MA, 1989. 39. McArthur JW, Bullen BA, Beitins IZ, et al: Hy- 24. Ellison PT: Morbidity,mortality, and menar- pothalamic amenorrhea in runners ofnormal che. Hum Biol 53:635, 1982. body composition. Endocr Res Commun 7:13, 1980. 25. Leistol K: Social conditions and menarcheal age: The importance of early years of life. Ann 40. Fraioli F, Cappa M, Fabbri A, et al: Lack of en- Hum Biol 9:521, 1982. dogenous opioid inhibitory tone on LH secre- tion in early puberty. Clin Endocrinol 20:299, 26. Malina RM: Growth and maturation: Normal 1984. variation and effect of training. In Gisolfi CV and Lamb DR (eds): Perspectives in Exercise 41. Frisch RE: Fatness of girls from menarche to Science and Sports Medicine, Vol 2. Youth, age 18 years, with a nomogram. Hum Biol Exercise, and Sport. Benchmark Press, Indi- 48:353, 1976. anapolis, IN, 1989, p 223. 42. Bronson FH,and Manning JM: The energetic 27. Drinkwater BL, Nilson K, Chestnut CH, et al: regulation of ovulation: a realistic role for Bone mineral of amenorrheic and eumenor- body fat. Biol Reprod 44:945, 1991. rheic athletes. N Engçl J Med 311:277, 1984. 43. Kuipers H, and Keizer HA: Overtraining in elite athletes: Review and directions for the future. Sports Med 6:79, 1988.

9CHAPTER Menstruation and Menstrual Disorders MONA M. SHANGOLD, M.D. PREVALENCE OF MENSTRUAL Chronic Hormone Alterations with DYSFUNCTION AMONG Training ATHLETES CONSEQUENCES OF MENSTRUAL REVIEW OF MENSTRUAL DYSFUNCTION PHYSIOLOGY Luteal Phase Deficiency TYPES OF MENSTRUAL Anovulatory Oligomenorrhea DYSFUNCTION Hypoestrogenic Amenorrhea MENSTRUAL CYCLE CHANGES DIAGNOSTIC EVALUATION OF WITH EXERCISE AND TRAINING MENSTRUAL DYSFUNCTION IN ATHLETES Weight Loss and Thinness Physical and Emotional Stress TREATMENT OFMENSTRUAL Dietary Factors DYSFUNCTION IN ATHLETES HORMONAL CHANGES WITH EVALUATION AND TREATMENT EXERCISE AND TRAINING OF PRIMARY AMENORRHEA Acute Hormone Alterations with Exercise Increased participation of women in sports has led to greater awareness of the menstrual cycle alterations that frequently accompany exercise and training. This raised consciousness has inspired more scientists to investigate the etio- logic mechanisms responsible for such changes and has led many athletes to seek medical attention. Unfortunately, many other athletes still avoid physician consultation, usually because they fear they will be told to stop exercising. It is the responsibility of all physicians and other health professionals to advise exer- cising women about what is known regarding reproductive effects of exercise and to assist them in formulating therapeutic plans. PREVALENCE OF MENSTRUAL DYSFUNCTION AMONG ATHLETES Oligomenorrhea (infrequent menses) and amenorrhea (absent menses) are more prevalent among athletes (10% to 20%)1,2 than among the general popula- tion (5%) and are found more often in runners than in swimmers or cyclists3 (Fig. 152

Menstruation and Menstrual Disorders 153 Figure 9-1. The prevalence of amenorrhea in runners, or number of years of training.2,4Bachmann swimmers, and cyclists, relative to training mileage. and Kemmann5 have reported that the prev- (From Sanborn,3 with permission.) alences of oligomenorrhea and amenorrhea among college students are 11% and 3%, re- 9-1), Among competitive athletes, the prev- spectively. However, this population in- alence of amenorrhea has been reported to cludes some athletes, for whom exercise be as high as 50%.3However, the prevalence and trainingcontribute to the problem. The of menstrual dysfunction does not correlate prevalence of menstrual dysfunction among with average weekly mileage, running pace, college students is higher than that among the rest of the population because college students tend to experience more emotional stress than the general population and be- cause many college students have not un- dergone full maturation of the hypotha- lamic-pituitary-ovarian axis, making them more susceptible to menstrual disorders. It is worth mentioningthat the general popu- lation has previously been considered to be sedentary, but the rising numbers of exer- cising women will undoubtedly increase the percentage of exercising women in the gen- Figure 9-2. Percent menstrual change during training for women with regular menses before train- ing, irregular menses before training, and amenorrhea before training. Of those women who had reg- ular menses before training, 93% continued to have regular menses during training. (From Shangold,2 with permission.)

154 Developmental Phases eral population and may raise the preva- result from variations in the length of the fol- lence of menstrual dysfunction in this licular phase, or the time required for a fol- group. licle to enlarge and mature enough to un- dergo ovulation. Although it is tempting to presume that exercise itself is responsible for the higher Throughout the menstrual cycle, the hy- prevalence of amenorrhea associated with pothalamus secretes gonadotropin-releas- it, many factors change simultaneously dur- ing hormone (GnRH), which is also re- ing the course of an athletic training pro- ferred to as luteinizing hormone-releasing gram, makingit difficult to isolate causal fac- hormone (LH-RH) or luteinizinghormone- tors. The fact that amenorrheic runners releasing factor (LRF). This decapeptide is have a higher incidence of prior menstrual produced by cells in the arcuate nucleus of irregularity1'2 suggests that exercise alone the hypothalamus; it promotes synthesis, may not be responsible for menstrual dys- storage, releasability, and secretion of both function in many cases (Fig. 9-2). pituitary gonadotropins: follicle-stimulat- ing hormone (FSH) and luteinizing hor- REVIEW OF MENSTRUAL mone (LH). FSH promotes growth of the PHYSIOLOGY ovarian follicle and synthesis of estrogen from androgen precursors. LH stimulates A brief review of menstrual physiology ovarian androgen production, maintaining a follows, to facilitate the understanding of supply of androgens available for conver- readers from diverse backgrounds. It is nec- sion to estrogens. essary to be familiar with the basic hor- monal events of the menstrual cycle, in In a normal menstrual cycle, a woman pro- order to appreciate both the hormonal and duces estrogen all the time and produces menstrual alterations that accompany exer- significant progesterone only after ovula- cise and training. For more comprehensive tion. Blood estrogen levels vary greatly reviews, the reader is referred to other pub- throughout the cycle, being quite low during lications.6-8 the early follicular phase and quite high dur- ing the late follicular phase. It is the high es- A normal menstrual cycle (counting from trogen level in the late follicular phase that the beginningof one period to the beginning triggers ovulation. During the luteal phase, of the next period) lasts from 23 to 35 days. levels of both estrogen and progesterone An ovarian follicle is the structure that con- are high. tains an egg; a corpus luteum is what devel- ops from a follicle after the egg has been ex- Estrogen stimulates the endometrium pelled. The follicular phase is the portionof (the inner lining of the uterus) to proliferate; the ovarian cycle that extends from the first progesterone promotes maturation and sta- day of menstruation until ovulation; this bilization of an estrogen-stimulated endo- corresponds temporally with the prolifera- metrium. It is the decline in the concentra- tive phase of the endometrial cycle. The lu- tions of estrogen and progesterone near the teal phase of the ovarian cycle extends from end of the menstrual cycle that results in ovulation until the onset of the next men- menstruation, which is the desquamation of strual period; this corresponds temporally the endometrium (Fig. 9-3). with the secretory phase of the endometrial cycle. A normal luteal phase should ap- TYPES OF MENSTRUAL proach 14 days, while a normal follicular DYSFUNCTION phase may vary considerably in length. Thus, fluctuations in the length of the men- With any insultto a woman's reproductive strual cycle of a woman who ovulates usually system, menstrual disturbance probably fol- lows an orderly sequence of increasing

Menstruation and Menstrual Disorders 155 Figure 9-3. Hormonal events of the menstrual cycle, phases of the ovarian and endometrial cycles, and endometrial height throughout the menstrual cycle. severity: (1) luteal phase deficiency, (2) amenorrhea is likely to ensue. Manywomen euestrogenic anovulation, and (3) hypo- do not seek attention when menstrual dys- estrogenic amenorrhea. Thus, any condition function is mild or of recent onset and may that disturbs the delicate balance of care- have hypoestrogenic amenorrhea by the fully timed hormonal events needed for reg- time they first seek attention. Although pro- ular ovulation and menstruation usually gression of this sequence has not been doc- produces luteal phase deficiency first. If the umented in prospective studies, it is likely, condition continues, euestrogenic anovula- nevertheless, and provides a useful model tion will probably follow. If the condition for understanding menstrual dysfunction. continues even longer, hypoestrogenic

156 Developmental Phases MENSTRUAL CYCLE CHANGES Table 9-1. FACTORS TO WHICH AN WITH EXERCISE AND TRAINING ATHLETE IS OFTEN SUBJECTED DURING TRAINING The data collected from the surveys re- 1. Weight loss ported are derived from records of women 2. Low weight who recorded only their menstrual patterns. 3. Low body fat Most, but not all, women who bleed atreg- 4. Dietary alterations ular intervals have normal ovulatory and lu- 5. Nutritional inadequacy teal function. More accurate information 6. Physical stress about menstrual cyclicity can be derived 7. Emotional stress from basal body temperature records and 8. Acute hormone alterations hormonal measurements. By having 14 sub- 9. Chronic hormone alterations jects record their basal body temperatures to indicate that and when ovulation had oc- have demonstrated that amenorrheic run- curred, Prior and co-workers9 have shown in ners were thinner and had lost more weight 48 menstrual cycles that even among ath- after initiating regularrunning. letic women with apparently regular men- ses, approximately one third have anovula- Despite claims that women need a mini- tion, one third have luteal phase deficiency, mum amount of body fat in order tomaintain and one third have normal luteal function. regular menstrual cyclicity, this hypothesis This suggests that menstrual disturbance remains unproven and suspect. If such a among exercising women may be more per- minimum amount of fat must be exceeded, vasive than has been appreciated. the mechanism by which this functions also remains unproven. Adipose tissue produces In addition to the epidemiologic studies and retains estrogen, but the amount of es- that demonstrate a higher prevalence of oli- trogen contributed by adipose tissue is neg- gomenorrhea/amenorrhea among athletes ligible compared with the very large quan- than among sedentary women, several pro- tity produced by normal ovaries. Since spective investigations have demonstrated muscle tissue contains aromatizing enzymes changes in menstrual cyclicity inindividual too, and since athletic women tend to have women who trained. Each of these has stud- more muscle and less fat than sedentary ied a number of factors that vary during women, aromatizing capability should be training, any of which may contribute to comparable in both groups. Thus, the mech- menstrual cycle alteration. It is usually very anism by which thinness promotesmen- difficult to separate the many contributory strual dysfunction remains to be shown. variables that change simultaneouslyduring training, including body composition, phys- Following the original suggestion by ical and emotional stress, diet, and certain Frisch and McArthur10 that thinness caused hormone levels (Table 9-1). amenorrhea, many investigators have probed the relationships between thinness Weight Loss and Thinness and hormone production and metabolism. Previously it was shown that thin women Many women lose both weight and body metabolize most of their estradiol by 2-hy- fat when they begin to exercise regularly. droxylation, while obese women excrete Some attain and maintain very low levels of most estradiol after 16-hydroxylation.11 Re- weight and fat. Simple weight loss and thin- cently, Snow and her associates12 have ness may lead to amenorrhea, even in the shown that elite oarswomen who develop absence of exercise. Shangold and Levine2 menstrual dysfunction during training me- have reported that amenorrheic runners are tabolize a greater fraction of administered lighter than eumenorrheic (regularly men- [2-3H]estradiol by 2-hydroxylation than do struating) runners. Schwartz and associates1 sedentary controls or elite oarswomen who remain eumenorrheic during training.How

Menstruation and Menstrual Disorders 157 the resultant catecholestrogens affect men- percentage of the total caloric intake of strual function remains to be shown. amenorrheic runners compared with that of eumenorrheic runners and eumenorrheic Physical and Emotional Stress nonrunners. These amenorrheic runners Schwartz and colleagues1 have shown that consumed more total calories than the other groups, however, so that equal quantitiesof amenorrheic runners associate more stress protein were consumed by all three groups. with their exercise than do eumenorrheic Calabrese and colleagues14 have demon- runners. This supports the concept that the strated that professional and student ballet physical and emotional stress of both train- dancers consume fewer calories (1358 calo- ing and competition may be substantially ries) than the recommended dietary allow- greater than appreciated. Although regular ance (RDA) (2030 calories) established by exercise tends to relieve stress and anxiety, the National Research Council,15 a figure in- this action may be outweighed in busy tended for an \"average\" woman,weighing women who are determined to incorporate 58 kg and exercising very little or not at all. a specific quantity of exercise into their Although the mean daily protein intake by daily schedules. these dancers (47.4 g) fell slightly below the RDA for \"average-sized women\" (50 g), this Warren13 has demonstrated the complex- protein intake was adequate when based on ity and interrelationship of the factors con- the RDAof 0.8 g/kg15 and the subjects' mean tributing to the development of menstrual weight of 53.1 kg. Frisch and associates16 dysfunction in two ballet dancers (Fig. 9-4). have reported that a group of collegiate The dancer in the upper graph experienced women who began athletic training prior to no change in weight or body composition menarche consumed less fat (65 g) and pro- throughout the year in which she had three tein (71 g) than a group who began training menstrual periods, each during an interval after menarche (95 g of fat and 92 g of pro- of inactivity. The dancer in the lower graph tein), and that the former group also had developed regular menses when she gained higher incidences of oligomenorrhea and both weight and body fat, although she amenorrhea. Very low levels of fat intake are maintained her customary level of activity. difficult to attain, and such diets have been She continued menstruating regularly, de- associated with insidious negative calcium spite a loss of both weight and body fat that balance.17 Deficiencies of the fat-solublevi- occurred during an inactive vacation inter- tamins, which require fat for absorption, val. With no further loss of weight, she have never been reported in people con- ceased menstruating altogether when she suming low-fat diets, but such deficiencies resumed her customary level of activity. It is remain a theoretical hazard. Deuster and likely that stress levels are higher during in- her co-workers18 have described differences tervals of intensive dancing, compared with between the dietary intakes of eumenor- vacation intervals. Thus, activity, fat, rheic and amenorrheic runners, and they weight, and stress must be considered vari- have reported that many amenorrheic run- ables in the changes observed. ners consume less than the recommended dietary allowances of some nutrients. Pirke Dietary Factors and associates19 have described menstrual Many women who begin to exercise regu- dysfunction that developed in association with caloric restriction, especially in asso- larly alter their dietary patterns because ciation with a vegetarian diet. These inves- they become more concerned about health- tigators have demonstrated impairmentof ful living. Those who have been exercising episodic LH secretion during dieting.20 regularly for a long time often eat differently from nonathletes. Schwartz and co-workers1 Despite the suggestion that amenorrheic reported that protein constituted a smaller runners may consume inadequate choles-

158 Developmental Phases Figure 9-4. Rdaenlacteiorsn.s(hFirposmamWoanrgremn,e1n3 wseist,h exercise, weights, and calculated body fat values in two young ballet permission.) terol to produce sufficient estrogen, there progesterone during a normal luteal phase, remains no evidence that dietary choles- but the rest of the body can provideenough terol is necessary for hormone synthesis. cholesterol to serve as precursor for ade- The corpus luteum cannot make enough quate luteal progesterone production. cholesterol de novo to synthesize adequate Although there is little to prove that estro-

Menstruation and Menstrual Disorders 159 gen production is affected by these dietary tradiol,24 progesterone,24 and testosterone23 differences, there is evidence that estrogen rise during exercise and return to normal metabolism is altered. Longcope and co- within an hour or two after cessation of ex- workers21 have shown that the ingestion of a ercise. Exercise-associated increments in low-fat diet promotes the same pattern of es- ACTH, opioid peptides, melatonin, and cor- trogen metabolism observed in thin women: tisol are facilitated by training.25,26 Since increased production of catecholestrogens testosterone27 and cortisol28 increase also in (the result of 2-hydroxylation) and reduced anticipation of exercise, it is probable that production of estriol (theresult of 16-hy- psychologic factors contribute to the re- droxylation). ported changes as well. Rebar and co- workers29 have shown that dexamethasone Myerson and her associates22 have shown suppression abolishes all effects of exercise that the resting metabolic rate (RMR)of on adrenal and gonadal hormones, includ- amenorrheic runners is significantly less ing those in anticipation of exercise. De- than that of eumenorrheic runners, which is tailed review of the many studies of hor- significantly less than that of eumenorrheic monal changes during exercise sessions, sedentary controls. The RMR of the amen- ranging in duration from a few minutes to orrheic runners remained lower than that of the time required to complete a marathon, is each of the other two groups after adjust- beyond the scope of this chapter. For a more ment for body weight or for fat-free mass. Al- comprehensive review, readers are referred though the absolute caloric intake of the elsewhere.30-32 amenorrheic runners was less than that of the eumenorrheic runners and was similar Factors Influencing Hormone Levels to that of the sedentary controls, the differ- ences were not significant, probably due to Plasma hormone levels represent a bal- large intragroup variability and small sam- ance among production, metabolism, utili- ple size. The amenorrheic runners also had zation, clearance, and plasma volume, all of higher scores on the eating attitudes test which may change simultaneously during (EAT-26, modified),including two subscales exercise. Levels of many hormones also are and total score; this reflected a higher level affected by episodic secretion, diurnal vari- of aberrant dietary patterns in the amenor- ation, state of sleep or wakefulness, state of rheic group. Thus, a growing body of infor- feeding or fasting, dietary composition and mation has brought our attention to the role caloric adequacy, temperature, body weight of dietary intake as a contributing cause of and composition, emotional factors, and menstrual dysfunction among athletes. body position. The hormonal response to exercise in any person is often influenced by HORMONAL CHANGES WITH the person's fitness, which affects the rela- EXERCISE AND TRAINING tive workload of any given activity and, in some cases, alters hormonal responsive- Acute Hormone Alterations ness during exercise. Difficulty in control- with Exercise ling these many variables during any spe- cific investigation makes it even harder to Blood levels of several protein and steroid interpret the observed exercise-induced hormones increase transiently during con- changes in hormone levels. tinuous, aerobic exercise. The long-termef- fects of such repetitive, but brief, alterations Chronic Hormone Alterations remains unknown. Reported exercise-in- with Training duced changes in gonadotropin levels are inconsistent and have been confused by the Shangold and associates33 have observed pulsatile nature of gonadotropin release. one runner during 18 menstrual cycles in Circulating concentrations of prolactin,23 es- which she varied her weekly mileage. This

160 Developmental Phases woman had shortening of the luteal phase Figure 9-6. Midluteal phase plasma progesterone con- and lower progesterone levels in cycles of centrations obtained 3 to 7 days after change in cervical greater mileage (Figs. 9-5 and 9-6). Prior mucus (presumptive evidence of ovulation), comparing and colleagues34 have also reported luteal seven samples from three control cycles and seven phase deficiency in two runners during sev- samples from three training cycles. Bars indicate eral menstrual cycles of varying mileage. means plus or minus standard errors (p < 0.001). One of these two runners had a normal preg- (From Shangold,33 reproduced with permission of The nancy when she stopped running, suggest- American FertilitySociety.) ing that exercise-induced luteal phase defi- ciency is a reversible phenomenon. alence of menstrual dysfunction was high in both groups during intensive training, but Similarly, Frisch and associates35 ob- was much higher in the weight-loss group; served a long-distance swimmer prior to, 94% of them experienced menstrual distur- during, and after intensive training, with bances, compared with 75% of the weight- monitoring of basal body temperature rec- maintenance group. Of those who lost ords, as well as blood and urine hormone weight, 63% experienced abnormal luteal measurements. She developed a luteal function, as did 66% of the weight-mainte- phase defect, followed by an anovulatory nance group. All subjects regained normal cycle, during intensive training. Three menstrual cyclicity within 6 months of ter- months after completion of a long-distance mination of the study (and presumably of swim (the English Channel), she regained a training). As has been shown by Warren,13 normal, biphasic basal body temperature weight loss and exercise act synergistically pattern. This confirms that the menstrual in promoting menstrual dysfunction. How- cycle alterations associated with intensive ever, these data36 suggest that a compensa- training occur in swimming as well as in run- tory increase in caloric intake cannot ning. prevent exercise-induced menstrual dys- function in most cases. Menstrual and hormonal changes in two groups of untrained women have been stud- In the same investigation of training-in- ied prospectively.36 One group lost weight duced menstrual dysfunction,37 two types of during a running program of increasing luteal dysfunction were described: a short mileage, and the other group maintained luteal phase and an inadequate luteal phase. weight during the same program. The prev- The short luteal phase was marked by de- creased luteal phase length, while the inad- Figure 9-5. Relationship between mileage run during equate luteal phase was characterized by in- the first 6 days of the follicular phase and the length of sufficient progesterone secretion, measured the luteal phase, defined as the interval between the by the concentration in overnight urine day of change in cervical mucus and onset of the next collections. The significance of these menses, in 18 cycles, (y = 13.3 - 0.1 Ix; r = -0.81; p differences remains to be shown, but these < 0.001). Point (1,13) represents three values. (From Shangold,33 reproduced with permission of The Ameri- can Fertility Society.)

Menstruation and Menstrual Disorders 161 investigators have shown that menstrual serum cortisol to bolus administration of dysfunction can be induced in normal human corticotropin-releasing hormone women with intense training. It remains to (CRH), and to meals, among both eumenor- be shown, however, whether a critical level rheic and amenorrheic athletes compared of exertion must be exceeded, and why some to eumenorrheic sedentary controls. These women are predisposed to this type of dys- data suggest that the hypothalamic-pitu- function in response to training. If a critical itary-adrenal axes of athletic women are level exists, the level undoubtedly differs characterized by increased CRH stimula- among various women. tion, increased cortisol negative feedback, normal ACTH secretion, normal cortico- Russell and associates38,39 found similar troph responsiveness to cortisol-induced weights and body fat levels among athletic negative feedback, and decreased respon- and inactive women, but found a correlation siveness to ACTH. In an excellent review, among strenuous exercise, anovulatory oli- DeSouza and Metzger44 have suggested that gomenorrhea, and elevated levels of /3-en- the adrenal response may be blunted be- dorphins and catechol estrogens. Although cause the adrenal is functioning near capac- endogenous opiates are known to modulate ity at rest, unable to mount a greater re- pulsatile luteinizing hormone release in hu- sponse to stimulation. mans,40 it is unlikely that circulating levels of these peptides correspond to the brain lev- Boyden and associates45 have provided an els influencing hypothalamic secretion. important clue toward understanding the al- terations in menstrual function associated The fact that a generalized increase in with intensive exercise. They have shown \"stress\" hormones occurs with exercise and that GnRH-stimulated LHlevels in eumenor- endurance training has been confirmed by rheic women decrease with endurance Villanueva and colleagues,41 who demon- training (distance running). strated increased cortisol production in both eumenorrheic and amenorrheic run- Gumming and co-workers46 further en- ners. Although the amenorrheic runners hanced our understanding of these changes had higher levels of both serum cortisol and when they reported that eumenorrheic run- urinary cortisol, the differences between ners (at rest) have lower LH pulse fre- these two groups of runners were not statis- quency, LHpulse amplitude, and area under tically significant. the LH curve over 6 hours, compared with eumenorrheic sedentary women (Figs. 9-7 Loucks and her associates42 have demon- and 9-8). These investigators47 then found strated that both eumenorrheic and amen- that acute exercise reduces LH pulse fre- orrheic athletes have higher morning serum quency but does not change pulse amplitude cortisol levels than do eumenorrheic sed- or area under the 6-hour curve. These im- entary women, and that the serum cortisol portant findings suggest that acute exercise levels in the amenorrheic athletes remained has an inhibitory effect on LH pulsatile re- higher throughout the day compared to lease at the hypothalamic level in eumenor- those in the eumenorrheic sedentary rheic runners, perhaps contributing to the women. However, these three groups did observed alterations with training. not differ in plasma ACTH pulse frequency, pulse amplitude, or mean level during any Several recent studies have provided time interval, and also did not differ in even more information about LH pulsatile serum cortisol pulse frequency. The eume- patterns in athletes. Veldhuis and co- norrheic athletes had reduced serum corti- workers48 demonstrated reduced LH pulse sol pulse amplitude during the day. Other frequency and normal LH pulse am- investigators have also described mild hy- plitude in amenorrheic or severely oligo- percortisolism in amenorrheic runners.43 menorrheic runners compared to eu- Loucks and co-workers42 have also shown a menorrheic sedentary controls. These blunted response of plasma ACTH and investigators also reported normal or accen-

162 Developmental Phases in the amenorrheic athletes, compared to the eumenorrheic sedentary controls. These data suggest that exercise-induced menstrual dysfunction results from inhibi- tion of hypothalamic release of GnRH at the level of the hypothalamus or higher brain centers influencing hypothalamic function. CONSEQUENCES OF MENSTRUAL DYSFUNCTION Luteal Phase Deficiency The major adverse condition associated with luteal phase deficiency is infertility, and this association remains controversial. Preliminary findings suggesting that proges- terone deficiency may be linked to an in- creased breast cancer risk49 have not been confirmed. Prior and her associates50 have recently demonstrated that shortening the luteal phase correlates with loss of bone density. Figure 9-7. Serum LHlevels in samples obtained at 15- Anovulatory Oligomenorrhea minute intervals over 6 hours in six eumenorrheic run- Chronic anovulation is associated with ners (upper) and four sedentary controls (lower). The studies were performed in the early follicular phase of chronic, unopposed estrogen production, the menstrual cycle (days 3 to 6). (From Cumming,46 which leads to continuous endometrial with permission.) stimulation and, as a consequence, an in- creased risk of endometrial hyperplasia and tuated LH release and normal estradiol re- adenocarcinoma. Although this association lease in response to exogenous GnRH has been documented in women with poly- pulses. cystic ovary syndrome,51-54 it has never been reported in athletes. It remains un- Loucks and co-workers42 have shown re- known whether anovulatory athletes carry duced LH pulse frequency and increased LH the same, increased risk of developing en- pulse amplitude in eumenorrheic athletes dometrial hyperplasia and adenocarcinoma compared to eumenorrheic sedentary con- as nonathletes with chronic anovulation. trols; both the LH pulse frequency and am- Perhaps inadequate reporting or history- plitude of the amenorrheic athletes were taking, or both, has led to the absence of lower than those of the eumenorrheic ath- such reports (i.e., gynecologists may not letes. An exogenous GnRH bolus caused routinely elicit athletic histories, particu- blunted FSH release in the eumenorrheic larly when diagnosing cancer), or perhaps athletes and augmented FSHand LHrelease anovulatory athletic women do not maintain high enough estrogen levels long enough to induce hyperplasia or cancer. Until this question is answered, it seems reasonable to assume that the endometrium of the athlete

Menstruation and Menstrual Disorders 163 Figure 9-8. LH pulse frequency, pulse amplitude, and the area under the LH curve in eumenor- rheic runners and sedentary con- trols in the early follicular phase of the menstrual cycle. (*p < 0.05, **p < 0.01 on Mann-Whitney U test.) (From Cumming,46 with permission.) responds the same as that of the nonathlete production may be iron-deficient or anemic. to estrogen stimulation. Thus, an increased Either of these conditions can impair ath- risk of endometrial hyperplasia and adeno- letic performance, as can heavy bleeding carcinoma should be presumed until it is during training or competition. The preva- disproved. lence of heavy bleeding among athletes re- mains to be shown. As suggested earlier, it Recent studies have suggested that an- is possible that anovulatory athletic women ovulatory women may also be at increased do not maintain high enough estrogen levels risk of developing breast cancer.55 This pre- long enough to induce sufficient thickening liminary report requires further confirma- of the endometrial lining and consequent tion. This suggestion, too, has not described profuse bleeding. However, heavy, infre- the athletic habits of subjects. Thus, if quent bleeding episodes are common chronic anovulation leads to an increased among adolescents, even those who are ath- risk of breast carcinoma, it remains to be letes; it is probable that more mature ath- shown whether this increased risk includes letes are subject to the same risk. anovulatory athletes. Hypoestrogenic Amenorrhea Although Frisch and associates56 have re- Estrogen promotes beneficial effects on ported a lower prevalence of breast cancer among former college athletes compared calcium metabolism, lipid metabolism, and with former college nonathletes, this report urogenital epithelial maturation. Hypoestro- did not relate breast cancer prevalence to genic women lack these favorable effects. recent athletic participation. Thus, it re- Many reports have demonstrated that ath- mains to be demonstrated whether regular letes with hypoestrogenic amenorrhea have exercise has any effect on breast cancer risk. reduced bone density and increased riskof Prior and her colleagues50 have shown that musculoskeletal injury, compared with eu- anovulatory cycles are also associated with menorrheic athletes.57-63 loss of bone density. Cann and co-workers57 were the first to Chronic anovulation usually leads to in- bring this finding to our attention. They re- frequent, heavy bleeding at unpredictable ported that women with hypothalamic times. At best, this is an inconvenience, par- amenorrhea, in many cases associated with ticularly to competitive athletes, and at exercise, had lower vertebral bone density worst, it may require hospitalization to con- than several other groups of eumenorrheic trol blood loss. Between these extremes, women with chronic, unopposed estrogen

164 Developmental Phases and amenorrheic women, including those entary women. This suggested that exercise with hyperprolactinemia and premature is beneficial in increasing bone density, but ovarian failure. This surprising, incidental not as beneficial as a normal estrogen level. findingled several other investigators to the Unfortunately, differences in calciumintake same issue. It had been shown by others that between some of these groups introduced exercise has a beneficial effect on bone den- another variable, as occurred in the Drink- sity, as discussed in Chapter 5. In view of the water study.59 It remains difficult to separate higher prevalence of hypoestrogenic amen- estrogen, exercise, and calcium intake as orrhea among athletes, it became important variables in pinpointing causality in such to resolve whether exercise is beneficial studies. enough to compensate for an estrogen defi- ciency. It was demonstrated by Jones and associates61 that radial bone density re- Rigotti and colleagues58 reported that gresses in a linear fashion with increasing amenorrheic women with anorexia nervosa duration of amenorrhea, regardless of etiol- had lower radial bone density than eume- ogy, confirming that hypoestrogenic young norrheic controls and that those anorectics women lose bone density in the same pat- who reported a high physical activity level tern as that observed for postmenopausal had a greater bone density than those who women.65 were less active. This suggested that physi- cal activity offers some protection against Warren and co-workers62 have reported bone loss induced by estrogen deficiency. that ballet dancers have a higher prevalence of scoliosis and a greater incidence of frac- In a study by Drinkwater and co-work- tures with increasing menarcheal age. They ers,59 lower vertebral bone density was also found a higher incidence and longerdu- found in amenorrheic athletes than in eu- ration of secondary amenorrhea among menorrheic athletes. However,these groups dancers with stress fractures. These find- differed not only in their estrogen status but ings suggest that menarcheal delay and pro- also in their calcium intake. Although the longed intervals of hypoestrogenic amen- absolute values of calcium ingested by orrhea may predispose ballerinas to the groups were not significantly different, scoliosis and stress fractures. the amenorrheic group, but not the eume- norrheic group, consumed much less cal- The suggestion of increased susceptibility cium than the amount recommended for to musculoskeletal injuries among amenor- hypoestrogenic women. Since estrogen en- rheic athletes has been supported by the hances calcium absorption, hypoestrogenic work of Lloyd and colleagues.63 These au- women require an additional 500 mg of cal- thors reported that women who were in- cium daily, compared with that required by jured during their running program were euestrogenic women. (It is recommended more likely to have had absent or irregular that euestrogenic women consume 1000 mg menses, were less likely to have used oral of calcium daily and that hypoestrogenic contraceptives, and had been running for women consume 1500 mg daily.64) Thus, it is more years than those running women who unclear whether the lower bone density of were not injured. these amenorrheic athletes was caused by estrogen deficiency, calcium deficiency, or The increased risk of cardiovascular dis- both. ease that occurs after menopause results mostly from adverse changes in lipids, in- Marcus and colleagues60 also reported duced by estrogen deficiency. Most of the that eumenorrheic runners had greater ver- adverse effects of the hypoestrogenic state tebral bone density than eumenorrheic sed- on low-density lipoprotein cholesterol con- entary women, who had greater bone den- centrations tend to be offset by endurance sity than amenorrheic runners, who had training. In addition, most athletes have a greater bone density than amenorrheic sed- reduced risk of cardiovascular disease, com- pared with the general population. On the

Menstruation and Menstrual Disorders 165 other hand, exercise-induced hypoestro- population. However, because these tests genic amenorrhea can reverse the beneficial have not proved cost-effective for patients effects of strenuous exercise on plasma apo- in my practice who have only menstrual dys- lipoprotein concentrations.66 function, I no longer perform these tests routinely. Because estrogen leads to maturation of the urogenital epithelium, a deficiency Menstrual disturbances may be caused by causes thinning of the vaginal epithelium hyperprolactinemia, hypothyroidism, ovar- and increased susceptibility to atrophic ure- ian failure, hyperandrogenism, and preg- thritis and vaginitis. These uncomfortable nancy. To detect these conditions, it is nec- conditions are most common after meno- essary to measure the following: serum pause, probably because development of prolactin, thyrotropin (TSH), free thyrox- urogenital atrophy requires several years ine, follicle-stimulating hormone (FSH), lu- in the hypoestrogenic state. Since few ath- teinizing hormone (LH), dehydroepian- letes remain severely hypoestrogenic long drosterone sulfate (DHEAS), testosterone, enough to develop atrophic vaginitis, this and B-human chorionic gonadotropin (B- condition is relatively uncommon among HCG). I also measure serum estradiol, in athletes and can usually be treated easily order to determine whether the patient is when it occurs. hypoestrogenic. Hyperprolactinemia may result from a pituitary adenoma or micro- DIAGNOSTIC EVALUATION OF adenoma; it requires further evaluation and MENSTRUAL DYSFUNCTION IN specific treatment. If both FSH and LH are ATHLETES very low, the sella turcica should be as- sessed (probably by a lateral cone-down I believe that all oligomenorrheic and film), to detect a large hypothalamic or pi- amenorrheic athletes deserve the following: tuitary lesion. An elevated TSHlevel or a low (1) a thorough history, including detailed free-thyroxine level indicates hypothyroid- dietary intake; (2) a physical examination, ism, which also requires further evaluation including a pelvic examination; and (3) and specific treatment. Hyperandrogenism some blood tests (Table 9-2). The dietary may result from any of several etiologies, record should be reviewed by a trained nu- including polycystic ovarian syndrome, ad- tritionist. Although most athletes with men- renal hyperplasia, an ovarian tumor, an strual disturbances will be found to have no adrenal tumor, or drug abuse; hyperandro- serious conditions, it is impossible to deter- genism requires further evaluation and mine, without this assessment, whether the treatment. Although many women with hy- menstrual dysfunction is related to exercise perandrogenism will also have peripheral or to some serious pathologic condition. A signs of androgen excess, not all women do. complete blood count, measurement of elec- Some hyperandrogenic women develop trolytes and liver enzymes, and urinalysis menstrual dysfunction before acne, hirsut- are useful screening tests for the general ism, or other symptoms of androgen excess. Therefore, I believe it is worthwhile to mea- Table 9-2. INITIALDIAGNOSTIC sure DHEAS and testosterone in all women EVALUATION OF OLIGOMENORRHEA OR with menstrual dysfunction, regardless of AMENORRHEA whether other symptoms are present. Preg- 1. History,including dietary intake nancy, of course, requires further care. 2. Physical examination, includingpelvic examination Ovarian failure requires at least counseling 3. Prolactin, free thyroxine, TSH,FSH, LH, DHEAS, and possibly also further evaluation and treatment. In a patient younger than age 30, testosterone, B-HCG, estradiol ovarian failure warrants a blood karyotype 4. Progestin challenge test to detect the presence of a Y chromosome, which confers an increased risk of gonadal

166 Developmental Phases malignancy. In a patient older than age 30, tion to protect the endometrium adequately. no further evaluation is required. At the time This can be effected by one of the following of the initial evaluation, and after blood has regimens: (1) medroxyprogesterone acetate been drawn for the above determinations, 5 to 10mg daily for 10to 14consecutive days the patient may be given a prescription for a of every month; (2) oral contraceptive pills, 5- or 10-day course of medroxyprogesterone each containing 30 to 35 mg of ethinyl estra- acetate, to assess whether herendometrium diol and 0.15 to 1.0 mg of progestin; or (3) has been stimulated by endogenous estro- clomiphene citrate to induce ovulation gen. If she has no withdrawal bleeding, her (Table 9-3). Ovulation induction should be endometrium had not been stimulated and reserved for those women desiring preg- the rest of her body probably also lackssuf- nancy at the time of evaluation. The first two ficient estrogen. Direct measurement of choices are acceptable for women who do serum estradiol gives more accurate infor- not seek pregnancy now, regardless of mation, however, and is more useful in plan- whether they are sexually active. Although ning treatment. oral contraceptive pills obviously provide contraception, medroxyprogesterone ace- After evaluatingan athlete witholigomen- tate does not, and this regimen requires in- orrhea or amenorrhea in this manner, and dividuals to use barrier contraceptive meth- upon finding that all of these tests except the ods if they are sexually active. estradiol concentration are within normal limits, the athlete can be reassured that se- Hypoestrogenic amenorrheic women re- rious causes of menstrual dysfunction have quire hormone replacement, primarily for been ruled out. She should be counseled skeletal protection, but also for urogenital about potential risks that may result from protection. Such athletes should be treated the condition. Her serum estradiol concen- with one of the following treatment proto- tration may be helpful in planning treat- cols: (1) conjugated estrogens 0.625 to 0.9 ment. mg daily and medroxyprogesterone acetate 5 to 10 mg daily on days 1 to 12 of every cal- TREATMENT OF MENSTRUAL endar month; (2) transdermal estradiol 0.05 DYSFUNCTION IN ATHLETES to 0.10 mg daily and medroxyprogesterone acetate 5 to 10 mg daily on days 1 to 12of Even if no serious causative pathology is every calendar month; (3) oral contracep- detected duringthe hormonal evaluation for tive pills, each containing30 to 35 g of ethi- menstrual dysfunction, treatment usually is nyl estradiol and 0.15 to 1.0 mg of progestin; indicated to prevent serious resultant pa- or (4) clomiphene citrate or human meno- thology. pausal gonadotropins to induce ovulation (Table 9-4). Ovulation induction should be The association between luteal phase de- reserved for women desiring pregnancy at ficiencyand infertility is generally accepted, the time of evaluation. but the links between luteal phase inade- quacy and breast cancer and bone loss seem Oral contraceptive pills may be recom- preliminary at the present time. Until con- mended to any hypoestrogenic amenor- firming studies for the latter two conditions rheic athlete who does not desire pregnancy are available, treatment for only infertility is recommended. Thus, at the present time,lu- Table 9-3. TREATMENT OF teal phase deficiency requires no treatment EUESTROGENIC OLIGOMENORRHEA unless and until pregnancy is desired. 1. If not sexually active or using barrier contraception: As discussed, euestrogenic anovulatory monthly progestin therapy women are at increased risk of developing 2. If contraception needed or preferred: oral endometrial hyperplasia and should be treated with monthly progestin administra- contraceptives 3. If fertility desired: clomiphene citrate

Menstruation and Menstrual Disorders 167 Table 9-4. TREATMENT OF Table 9-5. ABSOLUTE HYPOESTROGENIC AMENORRHEA CONTRAINDICATIONS TO ESTROGEN 1. If fertility desired: clomiphene citrate THERAPY 2. If contraception needed or preferred: oral 1. Abnormal liver function 2. History of thromboembolic or vascular disease contraceptives 3. Breast or endometrial carcinoma 3. If contraception and fertility not of concern: cyclic 4. Undiagnosed vaginalbleeding estrogen and progestin therapy fluctuations in their observed responses, 4. If diet inadequate: correct deficiencies and many experience psychologic benefit 5. If very thin: weight gain? from the regularity and predictability of oral 6. If exercising very heavily: less exercise? contraceptive therapy. at the time of evaluation, regardless of The major advantages of taking either whether she is sexually active; no additional progestin alone or estrogen and progestin as contraceptive method is needed by athletes separate pills are the ingestion of more selecting this form of hormone replacement physiologic doses of medication and the therapy. Those who select the more physi- likelihood of having predictable bleeding. ologic regimen of conjugated estrogens or Although the risks of exogenous hormone transdermal estradiol and medroxyproges- administration are much less than the risks terone acetate, separately, should be ad- of hormone deficiency, in my view, certain vised to use mechanical methods of contra- women should probably avoid estrogen and ception if they are sexually active. The major others should definitely avoid it. Absolute advantages of taking oral contraceptive contraindications to estrogen therapy are agents are convenience and contraception; listed in Table 9-5; relative contraindica- the major disadvantages are their two most tions are listed in Table 9-6. common side effects: breakthrough bleed- ing (bleeding on the days of pill ingestion) Many athletes have an aversion to exoge- and amenorrhea (lack of withdrawal bleed- nous hormone ingestion and do not compre- ing at the end of the hormone-containing hend the difference between physiologic re- pills in each package). These side effects are placement and pharmacologic therapy. It inconvenient but not serious; both can be al- requires careful and concerned counseling leviated by hormone manipulation. The low- to convince many of these women that hor- dose oral contraceptive pills recommended mone replacement therapy is advisable. are associated with much lesser side effects and complications than the higher doses If the dietary intake record reveals caloric prescribed commonly more than a decade or other nutritional inadequacy, the athlete ago; the low-dose preparations are also as- should be evaluated and counseled by a nu- sociated with a reduction in many disease tritionist and possibly a psychologist or psy- risks, compared with the risk to the general chiatrist, if an eating disorder is suspected population. Table 9-6. RELATIVE Another advantage of oral contraceptives CONTRAINDICATIONS TO ESTROGEN for athletes with menstrual dysfunction is THERAPY predictable bleeding and continued endo- 1. Hypertension metrial and skeletal protection. Many ath- 2. Diabetes mellitus letes may produce enough endogenous 3. Fibrocystic disease of the breast estrogen to have withdrawal bleeding fol- 4. Uterine leiomyomata lowing progestin administration for several 5. Familial hyperlipidemia months and then produce too little estrogen 6. Migraine headaches to do so during the next few months. It is dis- 7. Gallbladder disease turbing to many athletes to experience such

168 Developmental Phases (see Chapter 17). Many athletes will be will- Table 9-7. RECOMMENDATIONS FOR ing to increase their food intake when they FOLLOW-UP OF ATHLETES WITH understand that dietary inadequacy may be OLIGOMENORRHEA OR AMENORRHEA contributing to the problem. Those who are 1. Annual history and physical examination unwilling to change their diets should be re- 2. Annual prolactin, TSH, free thyroxine, FSH, LH, ferred for such counseling by a specialist. DHEAS, testosterone, B-HCG, estradiol Although some of them may prefer to gain 3. Annual progestin challenge test weight or to reduce training intensity or 4. Hormone replacement therapy quantity, to see if menses return without hormone therapy, it is not recommended ate on the day of an important competitive that these measures postpone for longer event and neverwantingto menstruate at all. than 6 months the initiation of hormone re- It is likely that most would prefer to have placement. A shorter trial is reasonable, normal reproductive function, rather than particularly if the athlete herself makes this amenorrhea, even if many are unwilling to suggestion. I believe that the benefits of reg- admit this to themselves. ular exercise far outweigh these potential re- productive hazards, which can and should EVALUATION AND be evaluated and treated if they develop. TREATMENT OF PRIMARY AMENORRHEA Despite the demonstration by several in- vestigators that exercise-associated men- Primary amenorrhea refers to the condi- strual dysfunction is often a reversible phe- tion in which menstruation has never oc- nomenon, there is no evidence that it is curred. Secondary amenorrhea,to which we reversible in all cases, nor is there any have referred until now, refers to the con- method of predicting when normal function dition in which menstruation had occurred will return, if ever. It seems unlikely that in the past but subsequently has ceased.Be- chronic, unopposed estrogen stimulation of cause menarche is often delayed in athletic the uterus will cause hyperplasia or adeno- girls, as discussed thoroughly in Chapters 7 carcinoma in an athlete in less than one and 8, it is tempting to assume that menar- year. However,endometrial hyperplasia can cheal delay is related to exercise. However, develop within 6 months in postmenopausal this assumption is as dangerous as that for women being treated with unopposed estro- secondary amenorrhea. Serious pathologic gen.67,68 This raises my concerns about conditions can easily be missed if they are permitting any women with euestrogenic, not sought. anovulatory oligomenorrhea to remain un- treated. Similarly, bone loss takes place at Any girl who has not developed any sec- an accelerated rate as soon as a woman be- ondary sexual characteristics by the age of comes hypoestrogenic, and a significant 13 should be examined and possibly evalu- amount of bone will be lost within the first 3 ated further. The same should be done for years of hypoestrogenism. I believe that it is any girl who has not begun to menstruate by best to initiate hormone replacement ther- age 16. Physical findings will direct appro- apy by the time 6 months have passed, for priate testing for these problems. As shown both oligomenorrheicand amenorrheic ath- in Table 9-8, the diagnostic evaluation of letes. I also believe that pelvic examination primary amenorrhea is similar to that for and blood evaluation should be repeated an- secondary amenorrhea, except for the nually in all athletes with menstrual dys- greater emphasis in primary amenorrhea function, regardless of whether they are re- upon detection of a uterus. ceiving hormone replacement (Table 9-7). Müllerian agenesis (which includes the Many athletes claim that they prefer to be absence of the uterus) is the second most amenorrheic. However, there is an obvious difference between not wantingto menstru-

Menstruation and Menstrual Disorders 169 Table 9-8. DIAGNOSTICEVALUATION OF loss, and exercise act synergistically to pro- ATHLETES WITH PRIMARY AMENORRHEA mote hormone alterations in both women 1. History, includingdietary intake with regular menses and those without. 2. Physical examination, includingpelvic examination 3. Prolactin, free thyroxine, TSH, FSH,LH, DHEAS, Athletes are more likely than sedentary women and girls to experience menstrual testosterone, B-HCG, estradiol dysfunction and menarcheal delay. How- 4. Progestin challenge test ever, this greater susceptibility should not 5. If uterus not palpable on pelvic examination: discourage any athletes from exercising in- tensely or frequently. The benefits of regular sonogram exercise far outweigh this potential hazard. 6. If uterus absent: testosterone, karyotype 7. If FSH high: karyotype The increased susceptibility of athletes to menstrual dysfunction also should not lead common pathologic cause of primary amen- to the presumptive diagnosis of \"exercise- orrhea, second only to gonadal dysgenesis. induced\" until completion of a comprehen- If the presence of a uterus cannot be deter- sive hormonal evaluation to rule out all mined with certainty by pelvic examination, other pathologic causes. It must be empha- a pelvic sonogram should be performed. sized that the diagnosis of \"exercise-related The third most common pathologic causeof menstrual dysfunction\" can be made only primary amenorrhea is androgen insensitiv- by excluding all other etiologies. Any ity syndrome (testicular feminization). woman or girl experiencing one of these Thus, the absence of a uterus requires fur- problems should be evaluated and treated. ther testing to distinguish between these two entities. The blood testosterone con- REFERENCES centration should be measured, and a blood karyotype performed. Abnormal findings 1. Schwartz B, Cumming DC, Riordan E, et al: should be followed with appropriate testing, Exercise-associated amenorrhea: A distinct as indicated. entity? Am J Obstet Gynecol141:662,1981. However, the most common cause of pri- 2. Shangold MM, and Levine HS: The effect of mary amenorrhea, particularly among ath- marathon training upon menstrual function. letes, is constitutional delay. Ifexamination Am J Obstet Gynecol 143:862,1982. indicates good estrogen effect, the girl can be reassured that menarche is likely to 3. Sanborn CF, Martin BJ, and Wagner WW: Is occur soon spontaneously. Copious estro- athletic amenorrhea specific to runners? Am genic cervical mucus usually indicates that J Obstet Gynecol 143:859,1982. spontaneous menarche will occur within 6 to 12 months. Hormone replacement ther- 4. Wakat DK, Sweeney KA, and Rogol AD: Re- apy for euestrogenic or hypoestrogenic ath- productive system function in women cross- letes is optional between the ages of 16 and country runners. Med Sci Sports Exerc 18, in my view, but should not be postponed 14:263, 1982. beyond the age of 18 because of the risk of osteopenia. 5. Bachmann GA, and Kemmann E: Prevalence of oligomenorrhea and amenorrhea in a col- SUMMARY lege population. Am J Obstet Gynecol 144:98, 1982. The tremendous increase in research in this field has enhanced our understanding 6. Judd HL (guest ed): Reproductive endocri- of the pathophysiology of exercise-associ- nology. Clin Obstet Gynecol 21:15, 1978. ated menstrual dysfunction. We now realize that, in many cases, dietary factors, weight 7. Shangold MM: Menstrual irregularity in ath- letes: Basic principles, evaluation, and treat- ment. Can J Appl Sport Sci 7:68, 1982. 8. Speroff L, Glass RH, and Kase NG:Clinical Gy- necologic Endocrinology and Infertility, 4th Ed. Williams and Wilkins, Baltimore, 1989. 9. Prior JC, Cameron K, Ho Yuen B, et al: Men- strual cycle changes with marathon training: Anovulation and short luteal phase. Can J Appl Sports Sci 7:173, 1982. 10. Frisch RE, and McArthur JW: Menstrual cy-

170 Developmental Phases cles: Fatness as a determinant of minimum 25. Carr DB, Bullen BA,Skrinar GS,et al: Physical weight for height necessary for their mainte- conditioning facilitates the exercise-induced nance or onset. Science 185:949,1974. secretion of beta-endorphin and beta-lipo- 11. Fishman J, Boyar RM, and Hellman L: Influ- tropin in women. N Engl J Med 305:560,1981. ence of body weight on estradiol metabolism in young women. J Clin Endocrinol Metab 26. Carr DB, Reppert SM, Bullen B, et al: Plasma 41:989, 1975. melatonin increases during exercise in 12. Snow RC, Barbieri RL, and Frisch RE: Estro- women. J Clin Endocrinol Metab 53:224,1981. gen 2-hydroxylase oxidation and menstrual function among the oarswomen. J Clin En- 27. Gumming DC, and Rebar RW: Exercise and r docrinol Metab 69:369, 1989. productive function in women. Am J Ind Med 13. Warren MP: The effects of exercise on puber- 4:113,1983. tal progression and reproductive function in girls. J Clin Endocrinol Metab 51:1150, 1980. 28. Hartley LH, Mason JW, Hogan RP, et al:Mul- 14. Calabrese LH, Kirkendall DT, Floyd M, et al: tiple hormonal responses to prolonged ex- Menstrual abnormalities, nutritional pat- ercise in relation to physical training. J Appl terns, and body composition in female clas- Physiol 33:607,1972. sical ballet dancers. Phys Sportsmed 11(2):86, 1983. 29. Rebar RW, Bulow S, Stern B, et al: Patterns of 15. Recommended Dietary Allowances, 9th Ed. endocrine response to exercise in normal Washington, DC, National Research Council, and dexamethasone suppressed women. Food and NutritionBoard, NationalAcademy Sixty-fifth annual meeting, Endocrine Soci- of Sciences, 1980. ety, 1983, Abstract 464. 16. Frisch RE, Botz-Welbergen AV,McArthur JW, et al: Delayed menarche and amenorrhea of 30. Shangold MM: Exercise and the adult female: college athletes in relation to age of onset of Hormonal and endocrine effects. Exerc Sport training. JAMA 246:1559, 1981. Sci Rev 12:53,1984. 17. Godara R, Kaur AP, and Bhat CM: Effect of cel- lulose incorporation in a low fiber diet on 31. Cumming DC, and Rebar RW: Hormonal fecal excretion and serum levels of calcium, changes with acute exercise and with train- phosphorus, and iron in adolescent girls.Am ing in women. Sem Reprod Endocrinol 3:55, J Clin Nutr 34:1083,1981. 1985. 18. Deuster PA, Kyle SB, Moser PB, et al: Nutri- tional intakes and status of highly trained 32. Loucks AB, and Horvath SM: Athletic amen- amenorrheic and eumenorrheic women run- orrhea: A review. Med Sci Sports Exerc 17:56, ners. Fertil Steril 46:636,1986. 1985. 19. Pirke KM,Schweiger U, Laessle R, et al: Diet- ing influences the menstrual cycle: Vegetar- 33. Shangold MM, Freeman R, Thysen B, and ian versus nonvegetarian diet. Fertil Steril Gatz M: The relationship between long-dis- 46:1083,1986. tance running, plasma progesterone and lu- 20. Pirke KM, Schweiger U, Strowitzki T, et al: teal phase length. Fertil Steril 31:130,1979. Dieting causes menstrual irregularities in normal weight young women through impair- 34. Prior JC, Ho Yuen B, Clement P, et al: Revers- ment of episodic luteinizing hormone secre- ible luteal phase changes and infertility as- tion. Fertil Steril 51:263, 1989. sociated with marathon training. Lancet 21. Longcope C, Gorbach S, Goldin B, et al: The 2:269, 1982. effect of a low fat diet on estrogen metabo- lism. J Clin Endocrinol Metab 64:1246, 1987. 35. Frisch RE, Hall GM,Aoki TT, et al: Metabolic, 22. Myerson M, Gutin B, Warren MP, et al: Rest- endocrine, and reproductive changes of a ing metabolic rate and energy balance in woman channel swimmer. Metabolism amenorrheic and eumenorrheic runners. 33:1106,1984. Med Sci Sports Exerc 23:15, 1991. 23. Shangold MM, Gatz ML,and Thysen B:Acute 36. Bullen BA,Skrinar GS, Beitins IZ, et al: Induc- effects of exercise on plasma concentrations tion of menstrual disorders by strenuous ex- of prolactin and testosterone in recreational ercise in untrained women. N Engl J Med women runners. Fertil Steril 35:699,1981. 312:1349,1985. 24. Bonen A, Ling W, Maclntyre K, et al: Effects of exercise on the serum concentrations of FSH, 37. Beitins IZ, McArthur JW, Turnbull BA, et al: LH, progesterone and estradiol. Eur J Appl Exercise induces two types of human luteal Physiol 42:15, 1979. dysfunction: Confirmation ,by urinary free progesterone. J Clin Endocrinol Metab 72:1350, 1991. 38. Russell JB, Mitchell D, Musey PI, and Collins DC: The relationship of exercise to anovula- tory cycles in female athletes: Hormonal and physical characteristics. Obstet Gynecol 63:452, 1984. 39. Russell JB, Mitchell DE, Musey PI, and Col- lins DC: The role of beta-endorphins and cat- echol estrogens on the hypothalamic-pitu- itary axis in female athletes. Fertil Steril 42:690, 1984.

Menstruation and Menstrual Disorders 171 40. Ropert JF, Quigley ME, and Yen SSC: Endog- polycystic ovary syndrome associated with enous opiates modulate pulsatile luteinizing endometrial cancer. Acta Obstet Gynecol hormone release in humans. J Clin Endocri- Scand 64:387,1985. nolMetab 52:583,1981. 55. Gonzales ER: Chronic anovulation may in- crease post-menopausal breast cancer risk. 41. Villanueva AL, Schlosser C, Hopper B, et al: (Medical News), JAMA 249:445, 1983. Increased cortisol production in women run- 56. Frisch RE, Wyshak G, Albright NL, et al: ners. J Clin Endocrinol Metab 63:133, 1986. Lower prevalence of breast cancer and can- cers of the reproductive system among for- 42. Loucks AB, Mortola JF, Girton L,and YenSSC: mer college athletes compared to nonath- Alterations in the hypothalamic-pituitary- letes. Br J Cancer 52:885,1985. ovarian and the hypothalamic-pituitary-ad- 57. Cann CE, Martin MC, Genant HK, and Jaffe renal axes in athletic women. J Clin Endocri- RB: Decreased spinal mineral content in nol Metab 68:402, 1989. amenorrheic women. JAMA 251:626,1984. 58. Rigotti NA, Nussbaum SR, Herzog DB, and 43. Ding J-H, Sheckter CB, Drinkwater BL, et al: Neer RM: Osteoporosis in women with an- High serum cortisol levels in exercise-asso- orexia nervosa. N Engl J Med 311:1601,1984. ciated amenorrhea. Ann Int Med 108:530, 59. Drinkwater BL, Nilson K, Chesnut CH, et al: 1988. Bone mineral content of amenorrheic and eu- menorrheic athletes. N Engl J Med 311:277, 44. DeSouza MJ, and Metzger DA: Reproductive 1984. dysfunction in amenorrheic athletes and 60. Marcus R, Cann C, Madvig P, et al: Menstrual anorexic patients: A review. Med Sci Sports function and bone mass in elite women dis- Exerc 23:995,1991. tance runners. Ann Int Med 102:158,1985. 61. Jones KP, Ravnikar VA, Tulchinsky D, and 45. Boyden TW, Pamenter RW, Stanforth PR, et Schiff I: Comparison of bone density in amen- al: Impaired gonadotropin responses to go- orrheic women due to athletics, weight loss, nadotropin-releasing hormone stimulation and premature menopause. Obstet Gynecol in endurance-trained women. Fertil Steril 66:5,1985. 41:359, 1984. 62. Warren MP, Brooks-Gunn J, Hamilton LH, et al: Scoliosis and fractures in young ballet 46. Cumming DC,Vickovic MM, Wall SR, and Flu- dancers. N Engl J Med 314:1348, 1986. ker MR:Defects in pulsatile LH release in nor- 63. Lloyd T, Triantafyllou SJ, Baker ER, et al: mally menstruating runners. J Clin Endocri- Women athletes with menstrual irregularity nol Metabol 60:810,1985. have increased musculoskeletal injuries. Med Sci Sports Exerc 18:374,1986. 47. Cumming DC, Vickovic MM, Wall SR, et al: 64. Heaney RP, Recker RR, and Saville PD: Meno- The effect of acute exercise on pulsatile re- pausal changes in calcium balance perfor- lease of luteinizing hormone in women run- mance. J Lab Clin Med 92:953, 1978. ners. Am J Obstet Gynecol 153:482,1985. 65. Meema S and Meema HE: Menopausal bone loss and estrogen replacement. Isr J Med Sci 48. Veldhuis JD, Evans WS, Demers LM,et al:Al- 12:601, 1976. tered neuroendocrine regulation of gonado- 66. Lamon-Fava S, Fisher EC,Nelson ME,et al:Ef- tropin secretion in women distance runners. fect of exercise and menstrual cycle status on J Clin Endocrinol Metab 61:557, 1985. plasma lipids, low density lipoprotein parti- cle size, and apolipoproteins. J Clin Endocri- 49. Cowan LD, Gordis L, Tonascia JA, and Jones nol Metab 68:17,1989. GS: Breast cancer incidence in women with 67. Schiff I, Sela HK, Cramer D, et al: Endometrial history of progesterone deficiency. Am J Ep- hyperplasia in women on cyclic or continu- idemiol 114:209, 1981. ous estrogen regimens. Fertil Steril 39:79, 1982. 50. Prior JC, Vigna YM, Schechter MT, and Bur- 68. Gelfand M, and Ferenczy A: A prospective 1- gess AE:Spinal bone loss and ovulatory dis- year study of estrogen and progestin in post- turbances. N Engl J Med 323:1221, 1990. menopausal women: Effects on the endome- trium. Obstet Gynecol 74:398,1989. 51. Fechner RE and Kaufman RH: Endometrial adenocarcinoma in Stein-Leventhal syn- drome. Cancer34:444,1974. 52. Jafari K, Ghodratollah I, and Ruiz G: Endo- metrial adenocarcinoma and the Stein-Lev- enthal syndrome. Obstet Gynecol 51:97, 1978. 53. Coulam CB, Annegers JF, and Kranz JS: Chronic anovulation syndrome and associ- ated neoplasia. Obstet Gynecol 61:403, 1983. 54. Dennefors BL, Knutson F, Janson PO, et al: Ovarian steroid production in a woman with

10CHAPTER Pregnancy MARSHALL W. CARPENTER, M.D. PHYSIOLOGIC CHANGES OF ACUTE METABOLIC RESPONSE PREGNANCY TO EXERTION ACUTE PHYSIOLOGIC RESPONSE MATERNAL THERMOREGULATION TO EXERTION IN THE DURING EXERCISE NONPREGNANT STATE ACUTE EFFECTS OF MATERNAL ACUTE METABOLIC RESPONSE TO EXERTION ON THE FETUS EXERTION MATERNAL EXERCISE TRAINING EFFECT OF PREGNANCY ON THE EFFECTS ON FETAL GROWTH ACUTE PHYSIOLOGIC AND PERINATAL OUTCOME RESPONSE TO EXERTION RECOMMENDATIONS ABOUT EFFECT OF PREGNANCY ON THE RECREATIONAL EXERCISE Exertion and pregnancy are the two most profound normal alterations in mammalian physiology. Exertion causes acute changes in cardiac output, blood flowdistribution, oxygen uptake, fuel mobilization, and the endocrine responses that facilitate these changes. Chronic exercise stress (exertional training) alters resting cardiovascular and metabolic homeostasis, the circulatory response to exertion, and aerobic capacity. Pregnancy appears to induce a primary vasodi- latation with associated increases in cardiac output, oxygen carrying capacity, oxygen uptake, and pulmonary changes. Whereas many of the cardiovascular changes that characterize pregnancy at rest are similar to those seen in acute exertion, the endocrine and metabolic changes of pregnancy differ considerably from those seen with acute exertion. Acute maternal adaptation to exertion and to exercise training has recently received increased investigational attention. The effect of maternalacute and chronic exercise stress on fetal homeostasis and growth and the role of maternal nutrition remain only superficially understood in humans, based primarily on animal investigation. The limited physiologic and epidemiologic investigation available, however, form the foundation for the guidelines and counsel that can be offered to pregnant women. This chapter examines the effects of pregnancy on resting physiology, and its interaction with the effects of acute exertion. The impact of acute exertion on 172

Pregnancy 173 fetal homeostasis and the effect of exercise in a 17%increase in minute ventilation rel- training on pregnancy outcome are also ex- ative to oxygen uptake (the ventilatory plored. These observations will be related equivalent).14 This results in a fall in arterial to recommendations which may be offered Pco2 from 39 to 31 torr, which produces a to pregnant women in clinical circum- mild respiratory alkalosis, increasing pH to stances. 7.44. Increased total lung capacity and in- creased tidal volume account for most of the PHYSIOLOGIC CHANGES OF increase in minute ventilation, rather than PREGNANCY changes in respiratory frequency. The in- creased resting oxygen uptake observed in Cardiovascular changes begin early in pregnancy is an early phenomenon, half of pregnancy and are well established by the which occurs by 8 weeks and three quarters midtrimester, thereby anticipating later by 15weeks' gestation. However, resting ox- fetal/placental requirements for oxygen and ygen uptake remains proportional to body nutrition. Plasma volume increases 45% by weight, not changingfrom the antepartum to 30 to 34 weeks,1,2 with measurable changes postpartum state.15'16 by 8 weeks. Despite a dilutional anemia, red cell volumeincreases by 20% to 30%by mid- ACUTE PHYSIOLOGIC pregnancy.2 RESPONSE TO EXERTION IN THE NONPREGNANCY STATE Cardiac output may increase secondary to a primary increase in circulating plasma vol- The cardiovascular and respiratory sys- ume or decreased systemic vascular resist- tems act in concert during acute exertion to ance,3 though the relationship of these fac- ensure adequate oxygen delivery to exercis- tors remains speculative.4 By 8 weeks, ing muscle while maintaining function in cardiac output increases by 23%and stroke other tissues. Oxygen consumption is the volume by 20%.5-7 The maximal increment product of oxygen delivery (heart rate, in cardiac output (34%) exceeds the 13% in- stroke volume) and oxygen extraction (ar- crease in body weight during pregnancy. teriovenous 0F2icdkifefqeureanticoen):,17as expressed in This is due, partly, to the 13% to 30% in- the modified crease in resting oxygen uptake observed in pregnancy8-10 and also to the decreased ar- Vo2 = HR-SV-avD02 teriovenous oxygen difference in preg- During incremental exercise to maximal nancy. intensity, Vo2 increases linearly to values typically 10to 20 times that at rest. Near the End diastolic volume6,7 and stroke peak intensity of exertion, a plateau of oxy- volume11 appear to increase through mid- gen uptake (Vo2) occurs, which persists de- pregnancy. Venouscompliance increases by spite greater exercise intensity. This upper the second trimester and is greater in limit of oxygen uptake (Vo2max) occurs as the lower extremities.12,13 These vascular maximal aerobic power is reached and is the changes and the expanding uterus may im- most important indicator of cardiovascular pede vena caval blood flow, so that maternal fitness. The percentage of Vo2max may be position increasingly alters measurements used, therefore, to describe relative inten- of hemodynamic function as pregnancy pro- sity of exertion among individuals with dif- gresses. The further increase in restingcar- ferent aerobic capacities when comparing diac output later in pregnancy seems to be physiologic responses that are related to ex- heart- rate dependent butvariable,due to dif- ertional intensity. Vo2max is usuallylimited ferences in maternal stature and position. by cardiac output. (See Chapters 1 and 4.) Cardiac output typicallyincreases four-to- Respiratory changes in pregnancy involve respiratory control and pulmonary function. Changes in respiratory control are reflected

174 Developmental Phases fivefoldfrom rest to maximum exertion. Car- oxygen uptake, approaching a ratio of 40 L of diac output increases with Vo2 in normal in- air per liter of O2uptake. This change in ven- dividuals in a ratio ranging from 5:1 to 6:1. tilatory pattern has been referred to as the \"ventilatory threshold\"21,22 but has uncer- Heart rate increases linearly with Vo2. Ini- tain physiologic significance.It is loosely as- tial increases during mild exertion result sociated with elevated levels of plasma lac- from release from vagal tone, and increases tate, found at high exertional intensity. at higher exercise intensities are caused by Exercise training results in a greater in- increases in sympathetic tone. Up to 40% crease in Vo2 at ventilatory threshold than in Vo2 max, stroke volume increases with in- Vo2max. Maximal voluntary ventilation does creased venous return to 1.5 to 2.0times that not limit Vo2max in normal individuals. at rest. Above a heart rate of 100, however, further increases in cardiac output are ACUTE METABOLIC RESPONSE pulse-dependent.18 TO EXERTION Peripheral as well as central hemody- The profound increase in the energy re- namic changes are necessary to effectively quirements of muscle duringexercise neces- deliver required oxygen and fuel to exercis- sitates the mobilization and distribution of ing muscle. Blood flow is redistributed by fuel from other tissues to sustain exertion sympathetic nerve activity, which is re- beyond the first seconds of movement. En- flected in increased plasma norepinephrine ergy consumption may increase over 10-fold concentrations.19 Norepinephrine concen- above resting values during intense exer- tration is closely related to intensity of ex- tion.21,23,24 Muscle can oxidize glucose, free ertion and to heart rate above 100.17 This fatty acids, glycerol and ketones to produce redistribution results in an early and sus- energy. The proportion of fuel types avail- tained linear reduction in splanchnic and able to muscle is a function of exercise inten- renal blood flow and, at high exertional in- sity, duration, nutritional state, and the tensity, causes decreased cutaneous perfu- physical fitness of the individual,and is de- sion. termined largely by the acute hormonal re- sponse to exertion. The proportion of total cardiac output perfusing exercising muscle increases with Carbohydrate stores in the body are the relative intensity of exertion regardless found in muscle glycogen (300 to 400 g, 5 • 103 of individual aerobic fitness. However, this kJ), hepatic glycogen (80 to 90 g, 1.5 •103 kJ), proportion is higher at maximal aerobic and blood glucose (20 g, 30 kJ).This is power among individualswith high levels of dwarfed by the energy stored as fat (about aerobic fitness. Therefore, the increment in 15 kg, 6-105 kJ). Protein is not significantly Vo2max obtained with exercise trainingis at- available as fuel during acute exercise. At tributable to increased oxygen uptake of ex- rest, free fatty acids provide the primary fuel ercising muscle, while nonexercisingvas- for muscle in the fasting state. cular beds receive the same low absolute blood flow. As exertional intensity increases beyond 60% Vo2max, carbohydrate is oxidized in Oxygen uptake during exertion is also en- higher proportions, so that at Vo2max, all the hanced by increased oxygen extraction from energy expended by muscle is derived from each volume of blood perfusing exercising carbohydrate oxidation. At this intense level muscle. This is reflected in a three- or four- of exertion, adenosine triphosphate (ATP) fold increase in arteriovenous oxygen differ- is provided increasingly by anaerobic gly- ence at maximal exertion compared to colysis, which is reflected in rising plasma rest.17'20 lactate concentrations above 60% Vo2max. Elevated plasma lactate may act to suppress Ventilation increases linearlywith oxygen uptake (at 20 to 25 L per liter of O2 uptake) to about 50% Vo2max, above which the in- crease in ventilation is greater, relative to

Pregnancy 175 lipolysis,25 thereby increasing demands on stimulate glucagon release, which, in turn, carbohydrate as fuel. Consequently, exer- augment hepatic glycogenolysis and periph- cise at Vo2max can only be sustained for a eral lipolysis.28,29 Both norepinephrine and short duration, being limited by the modest epinephrine increase with percent Vo2max stores of carbohydrate available to sustain and pulmonary artery oxygen saturation. exertion at this intensity. Most investigators Epinephrine concentration is increased employ some criterion for a \"plateau\" of ox- with intense exertion, is produced by the ad- ygen uptake with increasing workload to es- renal medulla, and correlates positively tablish that Vo2max has been achieved. The with norepinephrine and negatively with uncertainty about criteria to establish a glucose concentrations.30,31 Therefore, the maximum Vo2 plateau and the subject's dif- net effect of these changes is to augment and ficulty in maintenance of this level of exer- sustain the release of glucose. tion make observations under this condition Exertion also alters the metabolic effects problematic, especially in pregnancy. Data of insulin.The drop in insulin concentration from such studies thereby require some during acute exertion does not impede the judgment in their interpretation. marked rise in peripheral glucose uptake during exercise.32 Under these conditions, Exercise duration also influences fuel me- only an absolute lack of insulin causes a re- tabolism. The immediate, local sources of duction in glucose uptake (in the pancre- energy (ATP and phosphocreatine) provide atectomized dog), suggesting that insulin energy for the first 6 to 8 seconds of muscle may only have a permissive role in periph- contraction. Glycogenolysis and local lac- eral uptake during intense exertion.32Iso- tate production provide carbohydrate for 1 lated exertion increases insulin-mediated to 3 minutes of exertion at maximal aerobic glucose uptake (insulin sensitivity) and glu- exertion. Exercise beyond 5 to 10 minutes cose uptake at maximaleffective insulin con- becomes increasingly dependent on free centration (insulin responsiveness) up to 48 fatty acids. Moderate-intensity exertion for hours after exercise.33 40 minutes results in a fourfold rise in glu- Intense or prolonged moderate exertion is cose production by glycogenolysis and glu- required to produce a rise in circulating lev- coneogenesis to maintain plasma glucose els of glucagon,29,34 growth hormone,35 and concentration for tissue with obligate glu- cortisol. Growth hormone response corre- cose needs. This response is reduced 15% to lates with Vo2 and plasma lactate concentra- 60% by glucose infusion, and 67%by glucose tion.35 and insulin infusion.26 EFFECT OF PREGNANCY ON Diet antecedent to exercise may alter ex- THE ACUTE PHYSIOLOGIC ercise capacity at Vo2max. A high carbohy- RESPONSE TO EXERTION drate diet following intense exercise increases muscle glycogen stores. Low car- The impact of pregnancy on exercise re- bohydrate diets decrease muscle and he- sponse to submaximal and maximal exer- patic glycogen. Exercise capacity is in- cise differs. During pregnancy, we found that creased when muscle glycogen stores are absolute oxygen consumption (L-min^1) augmented.21'23,2427 Carbohydrate ingestion was 14% higher at rest, 9% higher during during exertion increases exercise endur- identical workloads during submaximal, ance.21'23,24.27 weight-supported cycle ergometry, and 12% higher during identical submaximal tread- The neuroendocrine response to exertion mill exertion16 compared to postpartum val- facilitates the mobilization of fuels for mus- ues. Similar investigations by others have cle contraction. Norepinephrine, released not consistently shown an increased oxygen from synaptic nerve endings, stimulates he- patic glycogenolysis and peripheral lipoly- sis. It also stimulates islet -adrenergic re- ceptors to inhibit insulin release and

176 Developmental Phases uptake during submaximal cycle exercise in Likewise, peak age-specific heart rate ap- pregnancy compared to nonpregnant con- pears to be unchanged by pregnancy at max- trols.8-11,37,38 However, identical submaximal imal aerobic exertion. Also, we and others treadmill exertion has been found to result have found no difference in maximal aerobic in increased Vo2 during pregnancy.8,16 When power in pregnancy when compared with oxygen consumption is expressed relative postpartum values.10 to body weight (mL-kg-2-min-1), however, there are no differences in submaximal ox- Recovery from exertion in the upright po- ygen uptake with either mode of exercise. Of sition may differ in pregnancy. Stroke vol- the increased oxygen uptake of submaximal ume recorded within 3 minutes of exercise exertion, 75% can be accounted for when the cessation has been observed to fall 26% dur- contribution of increased maternal weight is ing the third trimester, compared to only controlled for experimentally during preg- 11%postpartum.40 Cardiac output did not dif- nancy. This was accomplished by compar- fer, being maintained by a compensatory ing non-weight-bearing and weight-bearing increase in heart rate. This change in post- exercise during pregnancy and postpartum, exertional recovery may be related to in- and by using weight belts during postpar- creased venous compliance and capacity tum weight-bearing exertion to mimic preg- and possible vena caval obstruction that nancy weight.16 These data suggest that may characterize late pregnancy, though gravid women have an increased resting and this remains to be documented. exertional percent Vo2max due, largely, to the increased metabolic demands of the EFFECT OF PREGNANCY ON conceptus as well as the increased workof THE ACUTE METABOLIC moving a heavier body. RESPONSE TO EXERTION Pregnancy may alter the relative contri- bution of stroke volume and heart rate to in- Pregnancy produces alterations in hor- creased cardiac output during incremental monal and metabolic homeostasis, which workloads at high levels of exercise inten- distinguish it from the nonpregnant resting sity. In the nonpregnant state, stroke vol- and exercising state. Pregnancy produces ume does not increase with incremental ex- insulin resistance, which is reflected in an ertion above 60%Vo2max;further increases elevated fasting insulin-glucose ratio. This in cardiac output are due to increased heart may be observed more quantitatively by in- rate. In contrast, limited data in pregnancy ducing hyperinsulinemia by intravenous suggest that further increases in stroke vol- infusion and measuring the rate of glucose ume are still possible above this level of ex- infusion required to maintain steady-state ertional intensity.10 This change in the rela- euglycemia. This euglycemia, hyperinsulin- tive contribution of stroke volume to emic clamp technique demonstrates a incremental cardiac output with increased reduced requirement for infused glucose workload alters the regression equation of during pregnancy in order to maintain eu- Vo2 on heart rate during pregnancy. Conse- glycemia compared to that required in the quently, the mathematic model for predict- nonpregnant state under the same hyperin- ing Vo2max from wsuobmmeanxiims aallteVroed2/.h39eart-rate sulinemic conditions.41 Insulin binding on data in pregnant red cells is unaffected by pregnancy, but In contrast to submaximal exertion, the binding is reduced on adipocytes during limited studies performed on pregnant pregnancy.42 Pregnancy is characterized by women at maximal aerobic exertion show postprandial hyperglycemia and by fasting little, if any, change in maternal cardiovas- hypoglycemia. Free fatty acid and triglycer- cular response under this condition. Preg- ide concentrations are increased in preg- nancy is not associated with any change in nancy. the usual coupling of Vo2to cardiac output.10 Acute hormonal responses to exercise