The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK Introduction Since women’s first involvement in the Olympic Games in 1900, great progress has been made in increasing the participation rates of women across most events. In 2000, women represented 44% of the competitors at the Sydney Summer Games, although this was not matched by coverage of women’s sports in the major newspapers of Belgium, Denmark, France and Italy. In these countries, women’s sport at the Sydney Olympics represented only 29.3% of the articles and 38% of photographs (Capranica et al. 2005). The number of published research studies on females to support the increased participation, can be reflected in the number of publications in the Journal of Applied Physiology in the first 5 months of 2006. This data suggests that there is still a short fall in the proportions of female studies when compared to those on male. Of those articles specifically referring to exercise in humans (n=36), 61 % were on males only, 5 % were on females only, 8 % were on males and females with the genders being studied separately and 26 % were on males and females, with no acknowledgment of the mixed gender. Collating these facts, together with the observation that many of the early findings on females and exercise have been found to be invalid because of poorly-controlled studies, it is clear that there is still much to be done. The aim should be to ensure that training sessions are optimised for performance enhancement in the female population and that training and performance are not detrimental to the longterm health of female participants. There are a number of different models of investigation. Studies which try to directly compare genders have difficulties in matching subjects and minimising the confounding variables inherent in cross-sectional studies. Tarnopolsky and Saris (2001) recommend that subjects be matched on both training history and maximal . . oxygen uptake ( VO2max). This permits account to be taken of the largely genetic ( V . O2max) and the environmental (training state) factors contributing to VO2max. If, in addition, data is expressed as lean body mass, it accounts for the differences in body fat between the genders. Although this has allowed a focus on those differences that are truly the effect of gender, there are still possible errors in expressing the exercise . intensity as a percentage of VO2max particularly when the percentage used is using the classical endurance intensity of 70%. Recently, studies (including those on a single sex) have exercised subjects at an intensity that expresses the intensity as a percentage . of lactate threshold (LT). The LT for trained subjects, lies around 80% VO2max but . can vary from below 60 % to above 90 % VO2max. Therefore even if subjects are . matched for training history, VO2max and corrected for body weight, but have not been matched for LT, the results can still falsely signify gender differences where none exist. An alternative approach, avoiding the need for cross-sectional studies, is based on the assumption that gender differences are the result of the sex steroids. In this way, it is

1

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK possible to investigate exercise responses during the natural fluctuation of these hormones during the female menstrual cycle. However, even in studies where methodology is tightly controlled the physiological variability of the sex steroids affords large inter and intra-variability with oestradiol concentrations during midluteal phase ranging from 82-232 pg.mL-1 and progesterone ranging from 8-40 ng.mL1 .This has led, from a mechanistic perspective, as well as a practical perspective, to the investigation of the influence of exogenous female steroids on physiology and performance. Administration of the oral contraceptive (OC) pill offers a model for the investigation of the female sex steroids and it can be mapped over a longitudinal time period, with individuals acting as their own controls. Casazza et al. (2002) suggests that administration should be for longer than 1 month to ensure that any chronic or cumulative effects of the steroids are not missed. OCs suppress normal menstrual cycle levels of oestradiol and progesterone by inhibiting the pituitary release of gonadotrophins and replacing the sex steroids with consistent levels of oestradiol and progesterone, thus limiting the variation in circulating levels. A multiphase OC pill contains a changing dose of both hormones throughout 3 phases of the cycle with 21 days of administration (typically ~ 35 µg of oestradiol with increasing doses of progestagen from 200-250 mg) and 7 days of non-administration thus closely mimicking the “natural” cycle of sex hormones. The monophonic OC pill provides a constant level of oestradiol and progestagen for 21 days of a relatively low dose (~30 µg oestradiol and 150 µg of a progestagen). Theoretically another model of investigation could be the administration of testosterone to females except ethics would only permit this under clinical conditions hence the validity of extrapolating the data to healthy young women is limited. Finally, extrapolation of data from animal studies is difficult because of the significantly higher oestrogen and progesterone levels of animals (Jacobs et al. 2005). Although it is clear there are still difficulties in the analysis of human data on the effects of the sex steroids on physiology and performance there has been increasing clarity in recent years over the reverse of this, that is the effect of training and performance on menstrual status and the interplay of menstrual status, dietary intake and osteoporosis referred to as the female triad (Yeager et al. 1993). This review of the female athlete, builds on existing reviews and texts in the area and relevant sections are divided into: gender differences, variation throughout the menstrual cycle and the effect of OCs. Performance Effect of gender The improvement in female running world records has increased faster than males between 1950 until the current time. However, the common prediction that women will run faster than men, particularly over longer endurance events, is believed not to be accurate. Recent and more relevant prediction models identify that endurance running world records are nearing their limit and consequently the gender difference of 8-14 % over distances from 1500-42,000 m is unlikely to decrease further. Even at

2

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK the longer distances, 100-200 km, there is no evidence for women covering the distance faster (Table 1). Although these gender differences are perceived as relatively small, Cheuvront et al. (2005), who provide a review of this data, point out that the models predict that women will only break the 4-minute-mile barrier in 2033 if at all, some 80 years after Roger Bannister. One reason for the differences in endurance performance is almost certainly related to body composition and the higher body fat content of women. However, even when this is taken into account and data are expressed, as lean body mass, differences still . exist. The critical component is the lower VO2max of females, most probably arising from a lower haemoglobin content. Sprinting too shows no indication that women can beat men. The principle difference in sprint performance is due to differences in muscle cross sectional area which is almost certainly due to the anabolic effect of testosterone. The evidence for testosterone as the anabolic agent derives from studies where testosterone or a synthetic analogue have been administered to healthy males or hypogonadal older men and in these instances re-utilisation of intracellular amino acids increases, leading to enhanced net protein balance. The effect of administration of ovarian hormones on protein synthesis is not clear, although animal and in vitro studies would suggest that oestrogen inhibits muscle protein synthesis (Tipton 2001). In addition to the greater muscle mass males have a greater percentage area occupied by Type 11 fibres and this is likely to contribute to the greater ground forces exerted by men although women’s greater Type I/II fibre area ratio benefit them during eccentric contractions with longer cross-bridge cycle (and attachment) times. Women are also more effective in combining eccentric with concentric contractions being able to re-utilise the energy absorbed in a preceding eccentric contraction more effectively than men (Sale 1999). As the difference in fibre composition is largely due to fibre size and not number, it is unclear if sprint and strength-trained women would still show larger Type 1 fibres if compared to endurance-trained men. It is more likely that there is a continuum of fibre size reflecting the dominant usage and this would be consistent with data reporting that skeletal muscle adaptations to resistance exercise training are similar in males and females with at least one study reporting no differences in myosin heavy chain or in sarcoplasmic proteins (Tipton 2001). In their review Cheuvront et al. (2005) speculate as to the future and propose that there still potentially remains one way in which women could reduce the performance gap and that is through the recent IOC medical commission recommendation which will allow athletes undergoing sexual re-assignment surgery to compete in their reassigned sex category whether re-assigned after or before puberty. The implication being that if the surgery occurs after puberty then these athletes could still have a significant advantage (i.e. more lean mass). Whether this inclusive policy is abused will be known only in the future. Although the gender differences offer material for an interesting debate, of more practical importance is whether the menstrual cycle and the administration of oral contraceptives affect performance.

3

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK Effect of the menstrual cycle There appears little support for the concept that menstrual phase affects performance, body composition, skeletal muscle contractile characteristics or cardiorespiratory . factors including VO2max , in spite of anecdotal evidence to the contrary (de Jonge 2003). An exception to this may be related to the recovery period between intermittent sprints where an improvement in work completed was evident in the luteal phase of the cycle supported by an increase in oxygen consumption during recovery (Middleton & Wenger 2005) although no other reports support this difference in work capacity with intermittent sprints. Effect of Oral Contraceptive Administration Using a longitudinal design, Casazza et al. (2002) have found that 4 months of triphasic OC administration causes significant increases in body mass, fat mass and an . 11% decrease in VO2max, which is greater than that reported for 10 months . administration of a monophasic OC, where no difference or little difference in VO2max has been found (Rickenlund et al. 2004). Anaerobic performance too appears to be negatively affected. When a triphasic OC is administered over 3 cycles it is possible to detect reduced anaerobic performance when the hormone milieu is represented by high oestrogen and high progesterone (pill day 16-18) as opposed to low levels of both hormones (pill day 26-28) (Redman &Weatherby 2004). The mechanism behind this may be the attenuation of the catecholamine response to exercise whilst taking OC. This will potentially reduce the vasoconstriction to inactive tissues and reduce the mean arterial pressure in the active muscle. In addition, elevated levels of nitric oxide, a potent vasodilator, have been found. Both of these responses to OCs could limit peak exercise performance (Casazza et al. 2002). During high intensity exercise it could be the effect that the high steroids have on glycogen sparing causing an increased dependence on lipid oxidation (Redman & Weatherby 2004). However direct evidence for the mechanism behind the reduced performance has yet to be elucidated. Energy Systems Effect of Gender Considerable attention has been paid to gender differences in fuel utilisation during exercise over the last 20 years. Although early problems with poorly controlled studies have now largely been addressed there are still equivocal results relating to gender differences. Fuels potentially oxidised by the muscle for energy are protein, fat and carbohydrate (CHO). Protein is of particular significance during prolonged exercise. Yet as of 2006, there are only 5 studies that have investigated gender differences in amino acid use during endurance exercise and one of the studies had very small numbers. The 4

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK conclusion of these studies is that the amino acid, leucine, has higher oxidation rates in males than in females and this may be related to males responsiveness to catecholamines. This effect was illustrated when ß blockers were administered during exercise. In men, this caused an up-regulation of leucine oxidation whereas women it increased their reliance on fat. This might suggest that nutritional supplementation of amino acids/ protein, may be better justified in men than women (Lamont 2005). Increased lipolysis, as indicated by a greater rate of glycerol appearance, is consistent in showing increased rates in females when compared to men during moderate exercise (Blaak 2001; Mittendorfer et al. 2002). An alternative source of lipid, intramuscular triacylglycerol (IMTG) has been more difficult to determine due to the inherent error involved in its measurement. However three studies (Mittendorfer et al. 2002; Roepstorff et al. 2002; Steffensen et al. 2002), the latter two both from the same laboratory, have now indicated an increased IMTG use in women. Conversely, a recent study (Zehnder et al. 2005), utilising magnetic resonance spectroscopy (MRS), which allows a direct, non-invasive measure of IMTG from exactly the same location before and after exercise, concluded that male athletes depleted more IMTG than females. Although it might have been thought that this latter methodology would yield a definitive answer the authors themselves acknowledge certain limitations including the possibility of a higher training status of the males compared to the females. This was reflected in the higher IMTG levels at rest in the males. As IMTG concentrations are directly associated with higher IMTG utilisation this may have confounded the data. Whether the increased lipolysis results in an increased whole body fat oxidation is unclear. Until 2001, studies concluded that during sub-maximal exercise the respiratory exchange ratio (RER) was lower in women than in men indicating women’s increased reliance on lipid as a fuel, when the intensity is around the optimal . for lipid utilisation (i.e. 45-65 % VO2max) (Tarnopolsky & Ruby 2001). However, subsequent studies using kinetic data have been unable to detect a difference (Mittendorfer et al. 2002; Roepstorff et al. 2002; Ruby et al. 1997) in whole body lipid utilisation. Stable isotopes, the necessary tools for kinetic studies are expensive and as a consequence these studies have relatively small numbers in addition to the subject matching problems already identified. A study recently published, although only using quantitative data from RER, investigated 157 men and 143 women and reported that the men had lower rates of fat oxidation and an earlier shift to using CHO as the dominant fuel (Venables et al. 2004). Evidence for the site of the differing carbohydrate (CHO) utilisation appears more robust, with biopsy, isotopic tracer and MRS studies concluding that muscle glycogen utilisation is greater for men than women. Tracer studies of glucose appearance however are more ambiguous with both a lower rate of appearance in women and no difference being reported (Tarnopolsky & Ruby 2001). With little consistency in the effect of gender on substrate metabolism there is little motivation to seek the underpinning evidence, although suggestions include the attenuated catecholamine response of females during sub maximal exercise. Classically enhanced catecholamines would be expected to enhance both lipolysis and

5

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK glycogenolysis, and this is unlikely to be the sole answer. Other potential causes include the influence of testosterone however administration to healthy active males had no influence on fuel utilisation pathways (Braun et al. 2005). Finally fibre compositional differences between the genders, with a relatively greater fibre area of Type I fibres in females, may also contribute to fuel selection. Until there is clarity as to whether gender differences do exist, these options remain unexplored. Effect of Menstrual Phase Only in the last few years, have stable isotopes been utilised to investigate differences in fuel use between follicular and luteal phases of the menstrual cycle. These studies have indicated no differences in glycerol (Casazza et al. 2004) and non-esterified fatty acids (NEFA) (Jacobs et al. 2005). An elegant experiment of Horton et al. (2006), using stable isotopes and human subjects, separated the effects of oestrogen and progesterone by measuring in 3 phases of the cycle; early follicular, representing low oestrogen and low progesterone, mid-follicular representing elevated oestrogen and low progesterone and mid-luteal representing elevated oestrogen and elevated progesterone. These authors hypothesised that NEFA and glycerol turnover would be greatest when oestrogen was elevated and progesterone low. This was based on the potential inhibitory effect of oestrogen on CHO utilisation and the anti-oestrogen affect of progesterone. However their findings were unable to confirm their hypothesis and they concluded that there was no affect on glycerol and NEFA turnover (as reflected by palmitate flux) of the menstrual cycle during rest and . moderate (50% VO2max) exercise. Effect of OC The administration of exogenous hormones through triphasic OCs does increase triacyglycerol mobilisation. A longitudinal study, associated the elevation in triacylglycerol mobilisation with elevated resting and exercise cortisol, a lipolytic hormone (Casazza et al. 2004). The same group (Jacobs et al. 2005), went on to identify that this increase in lipolytic rate was not matched by an increase in whole body or plasma NEFA oxidation either at rest or during moderate exercise. Using a novel method for the determination of plasma and local NEFA oxidation the authors concluded that OC use increases both plasma-derived and total NEFA reesterification. Investigation of glucose kinetics has used different models of investigation, but regardless of whether oestradiol is given to amenoerrheic women (Ruby et al. 1997), to males (Carter et al. 2001) or through OC administration (Suh et al. 2003) the conclusion is that glucose flux decreases. As there is no change in whole body CHO this infers an increased use of skeletal muscle glycogen or lactate to compensate. In summary, it would appear that gender comparisons are inherently difficult to interpret because of an inability to truly match subjects. From the studies that have attempted, it can be concluded that amino acid and glycogen utilisation is greater in men, whilst lipolysis is higher in women (Table 2). The effect of menstrual cycle on

6

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK fuel selection in humans would appear to be minimal, although on administration of OCs there is an increase in lipolysis. However this does not result in increased fat oxidation but an increase in re-esterification from both plasma and the local skeletal muscle. OCs will also decrease glucose flux, which may be compensated by an increase in skeletal muscle glycogen or lactate. Menstrual Cycle Disorders The fact that physical activity, accepted as being beneficial for young and adult females, could have a negative impact and cause menstrual cycle disturbances was eventually accepted by those involved in sport in the 1970’s. Since then, the menstrual cycle disorder has been linked to disordered eating and osteoporosis. These three conditions represent an interlinked disease, although each can occur in isolation. This linking of the three conditions has been termed the female triad (Yeager et al. 1993). The incidence of menstrual cycle disorders may be underestimated due to the fact that disorders can occur which are asymptomatic. These conditions include “luteal phase defects” where the luteal phase does not produce sufficient progesterone to allow implantation of the fertilised egg. A more severe from of this is when the follicular development is so impaired that no ovulation occurs but there is some proliferation of the uterine lining. In both of these conditions athletes may believe that they are menstruating “normally”. Even accounting for this potential underestimation, the evidence is compelling that the incidence of oligomenorrhoea (defined as menstrual cycles of longer than 35 days) reported at 61% in rhythmic gymnasts one year after menarche and at 21- 40 % in runners in the gynaecological age range 5–30 years is widespread. Secondary amenorrhoea (absence of menstrual cycle for 3 months) is variably reported ranging from 2-31% in long distance runners and appears to be distance and age related, with higher mileage and younger age range reporting higher figures. These figures are much higher than reported for age-matched groups in sedentary women (9% for oligomenorrhoea and 2-5% for amenorrhoea) (Redman & Loucks 2005). These menstrual cycle disorders are due to a disruption of the pulses of leutinising hormone (LH) in the blood. It is these pulses of LH that dictate the release of the ovarian hormones and is the basis of ovarian function. Many variables can affect the ovarian axis but for a number of years the higher incidence in athletes was linked to body fat levels with primary amenorrhoea (absence of menses in a girl by 15 years) being linked to a body fat of less than 17 % and secondary amenorrhoea similarly being linked to a critical value of 22 % (Frisch & McArthur 1974). This hypothesis has now been modified and the more recent literature demonstrates unambiguously that it is the influence of energy availability, which affects the ovarian axis in athletes. This can be linked quantitatively with 5 days of an energy intake at ~125 kJ.kgFFM1 .day-1 leading to disruptions of LH pulsatile activity (Redman & Loucks 2005) (Figure 1). Some amenorrhoeic athletes report intakes as low as 67 kJ.kgFFM-1.day-1 alongside exercise programmes (Loucks & Nattiv 2005). The trigger of food deprivation is currently thought to be plasma glucose and/or leptin although the specific mechanism has yet to be elucidated (Wade & Jones 2004). 7

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK The consequence of the disturbed LH is low circulating oestrogen, which profoundly affects bone mineral density. Numerous studies have consistently reported bone mineral densities in young athletes more normally associated with postmenopausal women (IOC Medical Commission 2005). In post-menopausal women, osteoporosis is associated with bone resorption pathology however in young athletes suffering from osteoporosis the pathology appears to be manifested in low rates of bone formation (Redman & Loucks 2005). Recognising this difference, the International Society for Clinical Densitometry (ISCD) in 2004 identified different criteria for the diagnosis of osteoporosis in young athletes as that of postmenopausal women. These criteria recommend that premenopausal women and adolescent girls who present with low bone densities, relative to their own age, and possess one or more risk factors for fracture (e.g. hypogonadism, eating disorders, previous fractures) are diagnosed as osteoporotic. Although there is no direct relationship between the number of cycles required per year and the maintenance of bone density, it is recommended that concerns be raised if an athlete has been amenorrhoeic for more than 6 months or has had numerous stress fractures. Specific clinical criteria for bone density measurement are presented in the IOC Position Stand (IOC Medical Commission 2005). A problem, which may exist in the diagnosis, is that exercise itself may mask osteoporosis in certain bones. For example, runners may have high bone density in the calcaneous but have low bone densities in the hip and spine whilst rowers can have average bone densities in the forearm but not in the hip. This suggests that the higher density of the peripheral limbs is sport specific (Nattrass & Drinkwater 1998) and should be considered in the estimation of the bone densities. An additional consequence of the low oestrogen in these athletes is the effect that this has on endothelial vasodilation. In amenorrhoeic athletes, perfusion of skeletal muscle is reduced and oxidative metabolism is impaired (Harber et al. 1998) thus potentially affecting performance and recovery in these athletes. Therapeutic intervention is recommended within the first year when the bone loss is most rapid and should initially focus on a strategy to make changes in nutrition and training (IOC Medical Commission 2005). In order to avoid these clinical consequences focus must be directed to ensuring that athletes increase their energy intake to match any increase in energy expenditure. This mismatch of energy intake and energy expenditure can occur, at least acutely, in athletes within a range of body fat levels and is no longer peculiar to those only with low body fat (Figure 2 close to here). The motivation not to increase energy intake occurs predominantly in sports where body image or leanness is perceived as, or is, of benefit. Eating disorders however represents a continuum with minimal levels of eating disorders being sub-clinical. The IOC Medical Commission (2005) recommend that if it is suspected that an athlete is suffering from an eating disorder then they should be referred to a dietician and if positively diagnosed should be considered by the coaching team to be “injured” and to receive treatment by a multidisciplinary team to initially enhance nutritional practices. Clear guidance on treatment regimens is provided in the IOC position stand (IOC Medical Commission 2005). If the athlete refuses treatment, training and competition should be withheld until they agree to

8

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK participate. This to a competitive athlete who overtly feels well is extremely harsh but reflects the seriousness of the condition. It is clear that early diagnosis is imperative and studies are now investigating whether it is possible to identify those athletes who are most susceptible. One such report which compared the susceptibility of elite athletes with non-athletes (Torstveit & Sundgot-Borgen 2005) identified some risk factors that make non-athletes more susceptible to the female triad than elite athletes. This report has led to an exchange of correspondence (Di-Pietro & Stachenfield 2005), which perhaps can be summarised to say that significant work in the prevention of the athlete triad has yet to be conducted. Energy Balance When females increase their energy expenditure through an exercise programme, they report lower dietary intakes than expected for their high activity level. This observation has been explained by errors arising because of the methodology of the dietary intake measures. Some of these errors were overtaken in a recent short-term study (McLaughlin et al. 2006) where males and females were matched for percentage body fat and ad lib eating was measured when subjects were unaware that this was being monitored. This supported the early observations and indicated that differences did exist in the energy intake between genders in response to an 8-day exercise programme with no dietary compensation occurring in females whereas males increased their dietary intake to match the energy expenditure. This situation is not sustainable however, and in the long term some other mechanisms must come into play. In long-term training studies females in general do not loose body fat and it may be that this represents some overcompensation for short-term energy deficits as some reports even suggest an increase in body fat. This gender difference is supported by cross-sectional studies indicating that there is no relationship between physical activity and body composition in females although this exists in males (Westerterp 1999). Although body mass reduction with exercise is related to initial body mass even when subjects are matched for body fat this gender difference is still evident (Andersson et al. 1991) and may be related to the observation that with a similar percentage body fat women have approximately 50 % more fat cells than men and within each fat cell there may be a minimal “plateau” to which the fat cell can reduce (Krotkiewski et al. 1983). If females do not lose body mass and reports suggest that dietary intake is not increased, at least in the short term then they may compensate by reducing their energy expenditure outwith the formal exercise. A major component of this is resting metabolic rate (RMR). However this appears to be more dependent on the intensity of the exercise than any gender differences per se. Decreases in RMR are noted in response to too high an exercise intensity yet in order to get an increase it must be at . least 70 % VO2max. Other aspects of physical activity outwith formal exercise include daily tasks and although there are few long-term studies in humans, the available evidence suggests that compensation can occur in this component if the volume

9

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK and/or intensity of the formal exercise is too high. This may contribute differentially in each gender. Intuitively however this might be more of an individual response rather than a gender one (Westerterp 1999). Location of the subcutaneous adipocytes differs between the genders with the classical location of the adipose tissue in females within the gluteal-femoral region, whereas in men higher amounts lie around the abdominal area. This reflects differences in sensitivity to catecholamine stimulation, lipoprotein lipase and insulin activity of the adipocytes in the two populations. Both sexes are more sensitive to catecholamines around the abdominal tissue but the difference is greater in females. This differing regional sensitivity can be illustrated when epinephrine is infused into subjects and the leg NEFA release is doubled in men but does not change in women. Conversely the release was higher in women than men from the upper body region (Blaak 2001). Studies on obese men and women give some credence to the concept that the location of fat influences fat loss with exercise. It is possible to divide obese females into those with “apple-shaped” obesity (android) and “pear-shaped” obesity (gynoid). Android-shaped obesity females respond more like males in that they will tend to lose body fat with an exercise programme whereas the gynoid-shaped obesity females do not decrease their fat and in some instances increase body fat with no change in lean body mass after the physical training (Krotkiewski & Bjorntorp 1986). However direct evidence is required before these regional differences in fat utilisation can be linked to differing energy balance responses to exercise in athletes. Menstrual Cycle Phase and OC Administration Studies have consistently shown, from dietary records, that energy intake is increased during the luteal phase of the menstrual cycle and that this may relate to the need to maintain energy balance as resting energy expenditure increases in parallel. These changes in energy balance are not evident when the menstrual cycle is suppressed with administration of depot medroxyprogesterone nor is there any change in body weight (Pelkman et al. 2001). The lack of gain in body weight tends to be a consistent finding whether it is a monophasic or triphasic OC. This finding is contrasted by Casazza et al. (2002) who reported increases in body fat and weight with a triphasic OC over 4 cycles suggesting a shift in the energy balance equation. Nutrition Although prevalence of under-reporting may question some of the reported data there are some athletes who undoubtedly have low energy intakes. Athletes participating in sports where low body weight is an advantage to performance are constantly concerned with dietary restrictions, which can lead to a negative energy balance which will negate any positive outcomes for performance and also lead to negative consequences of inadequate nutrient intake, a higher risk of injury and increased risk of illness. Under weight maintenance conditions, the athletic female must meet sufficient kJ through intake to meet expenditure. Although energy intakes of less than 7,500kJ are unlikely to meet the energy requirements of athletes, theoretically women, who eat more than 5000-6500 kJ, with a variety of food sorts, can obtain most of the nutrients 10

The Female Athlete (2009). The Olympic Textbook of Science and Sport, The

Encyclopaedia of Sports Medicine An IOC Medical Commission Publication. Chapter 23 382 -397 Wiley- Blackwell , Chichester, UK with the possible exception of iron. Women will loose on average 1.3 mg of iron daily during menstrual flow which is double that of men and may lead to women being more susceptible to iron deficiency anaemia. Iron depletion occurs in 3 stages: iron depletion, iron deficient erythropoeisis and iron deficient anaemia. Iron depletion is identified through serum ferritin levels ( Female Male > Female Female > Male Male

24