Metabolic effects

As shown in Table 1 (see supplementary material), Andersson et al. (2016) reported on 13 highly trained (66 ± 1 mL·min⁻¹·kg⁻¹) men (n = 9) and women (n = 4) who performed submaximal incremental cycling tests (3 min each at 30%, 40%, 50%, 60%, 70% and 80% of maximal oxygen uptake) in either a fasted or fed state. Maximal fat oxidation rates were higher with fasted exercise (fed: 0.51 ± 0.04 g·min^−1, fasting: 0.69 ± 0.04 g·min⁻¹, p < 0.01). Due to the small number of women included in the study, the study lacked sufficient power to detect sex differences^{(Reference Andersson Hall, Edin and Pedersen8)}. However, whilst the authors noted a tendency towards lower maximal fat oxidation rate and higher Fatmax, the exercise intensity at which the greatest amount of fat is utilised per hour, for women compared with men, the overall pattern of changes in fat oxidation was comparable for both sexes. While this study suggests that men and women respond similarly in fed and fasted states, the small sample size and limited statistical power to detect differences limit our conclusions. A second study investigated the effects of low-volume high-intensity interval training (HIIT) performed in the fasted versus fed state on body composition, muscle oxidative capacity and glycaemic control in 16 women with overweight/obese^{(Reference Gillen, Percival and Ludzki9)}. Following 18 sessions of HIIT (10 × 60-s cycling efforts at 90% maximal heart rate, with 60-s recovery) for 6 weeks, no differences were reported in any of the measured variables between the fed and fasted states. Although both groups showed changes in body composition during the intervention, fasting versus feeding did not affect these changes. The applicability of these findings is limited by the population group (overweight/obese) and may not reflect an active female population, as well as by the study’s small sample size. The evidence is sparse, but to date, research does not suggest sex-based differences in any insulin, carbohydrate (CHO) or fatty acid markers, nor are there differences between women exercising in a fed or fasted state^{(Reference Gillen, Percival and Ludzki9)}.

A well-cited study promoting sex differences in response to fasted versus fed training is based on data from mouse models^{(Reference Piotrowska, Tarnowski and Zgutka10)}. Whilst rodent studies have a role in establishing potential mechanistic links, they do not provide evidence for direct application to humans. Further, the aforementioned study does not report differences in exercise between fed and fasted conditions; instead, it compared the effects of prolonged fasting (every-other-day fasting) for nine months on liver proliferation and apoptosis. The livers of male mice stopped producing energy storage molecules with fasting, whereas the livers of the females utilised all energy stores, including muscle, to maintain reproductive capacity. Whilst they observed an increase in the lipid content of the livers of every-other-day-fed female mice, there was also a decrease in proliferation and apoptosis in the livers of both female and male mice fed every other day, suggesting that tissue maintenance occurred during this trial^{(Reference Piotrowska, Tarnowski and Zgutka10)}. Whether this translates to human females is difficult to conclude, especially when human research does not agree.

Body composition

Vieira et al.^{(Reference Vieira, Blanco-Rambo and Bandeira-Guimaraes11)} investigated the effects of a 12-week resistance training programme on muscle size and strength in women ( n = 22) and men (n = 6). Due to the high proportion of women in the study, the results were analysed as a whole sample and separately for women. They showed improvements in muscle thickness, maximal dynamic strength (the highest force a muscle group can generate during a single voluntary contraction involving movement), and muscle power over time, but no differences between fasted and fed states for the total sample, nor for women only. Again, suggesting differences between men and women do not exist, nor do differences between fed and fasted training, and, in fact, to the contrary, that training fasted with resistance exercise can be anabolic. Further, 43 trained women who completed 6 weeks of high-intensity resistance training (HIRT) showed no differences in strength nor body composition changes between the fed vs fasted groups, nor was a difference in resting energy expenditure observed, but the fed state did result in greater fat oxidation (assessed via RER) at 30 min post-exercise^{(Reference Pihoker, Peterjohn and Trexler12)}. Schoenfeld et al.^{(Reference Schoenfeld, Aragon and Wilborn13)} examined body composition changes during an energy deficit in 20 young women (mean age 22.4 ± 2.8 years) who trained in a fed or fasted state. Participants performed 1 hour of steady-state exercise 3 days per week for 4 weeks, receiving a meal replacement shake either immediately prior to exercise (fed) or immediately after (fasted). Although the intervention led to reductions in weight and fat mass, body composition changes did not differ between conditions. Overall, the results of studies to date suggest that there is no significant difference in body composition changes between training in a fasted or fed state; however, resistance training or creating an energy deficit by exercise and diet results in favourable body composition changes.

Hormones

Kisspeptin, a protein hormone, has been proposed to link perturbations in energy balance with alterations in the activity of the reproductive axis^{(Reference Clarke, Dhillo and Jayasena14)}. Differences in kisspeptin have been observed between males and females^{(Reference Hrabovszky, Ciofi and Vida15)}. Further, kisspeptin has been evidenced to play a role in fertility and functional hypothalamic amenorrhoea^{(Reference Podfigurna, Maciejewska-Jeske and Meczekalski16,Reference Xie, Kang and Zhang17)} . However, no study has measured kisspeptin levels following fasted training in healthy females. Fundamentally, whether this can be extrapolated to healthy female athletes training fasted, when otherwise in adequate energy availability, is contentious. Whilst there appear to be some similarities in kisspeptin between species, there are also a number of differences^{(Reference Goodman and Lehman18)}.

Cortisol is a stress hormone which fluctuates throughout the day and night and can elevate with exercise but not beyond normal physiological ranges, and with training this response is blunted^{(Reference Hackney and Walz19)}. Whilst cortisol has catabolic actions, it plays a beneficial role in recovery due to its anti-inflammatory properties and protein turnover^{(Reference Hackney and Walz19,Reference Viru and Viru20)} . Indeed, exercising in a fasted state or with an inadequate CHO diet for several days can augment the cortisol response to exercise, similarly this adaptation occurs when exercising in extreme environmental conditions^{(Reference Hackney and Walz19)}. Crucially, these cortisol elevations are temporary and within the normal physiological range^{(Reference Hackney and Walz19)}. No differences in salivary cortisol were observed between males (n = 8) and females (n = 8) following a 2 hour cycling test at 65% VO₂max (an exercise intensity that elicits 65% of the maximum volume of oxygen the body can consume and utilise) after an overnight fast in either the fed (2210 kJ) vs fasted state^{(Reference Allgrove, Geneen and Latif21)}, suggesting fasted exercise does not elicit a unique cortisol response in either sex. Therefore, it is unwarranted to conclude that fasted exercise elevates cortisol in a way that increases in abdominal fat or induces harmful levels of sympathetic nervous system activity in otherwise healthy exercising females. However, there is a caveat for an athlete in state of problematic low energy availability (LEA), then research does show cortisol, luteinising pulsility and thyroid function can be impacted by exposure to LEA^{(Reference Loucks, Verdun and Heath22)}. Those with Cushing’s syndrome or other health conditions which affect cortisol are also cautioned. It is entirely plausible that these groups may experience different effects from fasted exercise due to their chronically elevated cortisol levels^{(Reference Daly and Hackney23)}.

In conclusion, the totality of evidence to date indicates that healthy women who wish to train in a fasted state are unlikely to suffer significant adverse effects, provided they do not place themselves in a state of problematic LEA. As such, the claim that all healthy women should not exercise in a fasted state does not appear to be supported by the research evidence. However, this statement is caveated to the individual situation and goal-dependent. There are of course exceptions to consider, (1) Session focus – if the focus of the training session is performance, research and nutritional guidelines for men and women suggest that for most a fed state is required, with an emphasis on CHO, (2) energy availability status - if the athlete is at risk of problematic LEA, then research has shown that eating around training is key to achieving adequate EA and for their recovery, (3) multiple sessions - if training multiple times a day, energy requirements are increased and opportunity to eat is reduced, before exercise becomes an important fuelling window to utilise, unless training fasted is forming part of a well-planned, monitored periodised training programme.

Menstrual cycle and fluid balance

For the purposes of this review, difficulties with accurately determining the phase of the menstrual cycle in naturally menstruating women will be set aside. The reader is directed to the paper by Burden et al.^{(Reference Burden, Altini and Ferrer24)} for a full review on these challenges, and why the menstrual phase based on the number of days should not be assumed.

During the luteal phase of the menstrual cycle, there is an elevation of 0.18–0.56°C in core temperature, and this increase aligns with progesterone peaking during the mid-luteal (ML) phase^{(Reference Charkoudian, Hart and Barnes25,Reference Kolka and Stephenson26)} . Conversely, oestrogen therapy in post-menopausal women has been shown to reduce core body temperature and thresholds for sweating and vasodilation^{(Reference Tankersley, Nicholas and Deaver27)}. Therefore, the evidence to date suggests that oestrogen and progesterone have opposing effects on body temperature, which could potentially impact fluid balance. Further, during the luteal phase, aldosterone levels increase, accompanied by a downward shift in the osmotic threshold for adenosine vasopressin (AVP) release^{(Reference O’Donnell, Floras and Harvey28,Reference Stachenfeld, Silva and Keefe29)} . Aldosterone increases Na retention and potassium excretion, leading to fluid retention^{(Reference O’Donnell, Floras and Harvey28)}, and AVP promotes water retention^{(Reference Stachenfeld, Silva and Keefe29)}. The increases in these two hormones have been reported to delay the onset of sweating and vasodilation during exercise^{(Reference O’Donnell, Floras and Harvey28,Reference Christison, Gurney and Williamson-Reisdorph30)} . This fluid retention can result in small transient fluctuations in body mass ∼0.5 kg during the menstrual cycle^{(Reference Kanellakis, Skoufas and Simitsopoulou31)}.

To determine whether fluid balance differs across the menstrual cycle, 11 eumenorrheic women (classified by the gold standards^{(Reference Elliott-Sale, Minahan and de Jonge32)}) walked for two 180-min trials in a heat chamber (35°C and 30% relative humidity) during the early-follicular (EF) and ML phases^{(Reference Christison, Gurney and Williamson-Reisdorph30)}. As expected, there was a significant difference in aldosterone, estrodiol, progesterone and core temperature between the phases. However, there was no effect on physical strain index, rating of perceived exertion, sweat rate, haematocrit, Hb, change in plasma volume, urine volume, nor percentage dehydration between the phases. Neither was there a phase effect on VO₂, fat oxidation or CHO utilisation, see Table 2 (see supplementary material) for study details. Although baseline body mass did not differ significantly between the two phases, there was a trend towards a higher starting body weight in the ML phase (∼0.2 kg, p = 0.066), potentially reflecting phase-related differences in aldosterone, which increases renal reabsorption of Na and water. Further evidence for the absence of menstrual cycle phase differences in fluid balance comes from Freemas et al.^{(Reference Freemas, Goss and Ables33)}, who investigated the effects of physical work in the heat (33.8 ± 0.8°C, 54 ± 1% relative humidity) during the EF phase, compared with the late follicular (LF) and ML phases. Twelve healthy, eumenorrheic women completed three 4 h trials (EF, LF and ML phases) with each trial involving 30 min of treadmill walking per hour with ad libitum fluid access. Fluid intake, total urine output, sweat rate and percentage body mass change did not differ between phases. Collectively, these findings indicate that although fluctuations in female sex hormones across the menstrual cycle alter the osmotic operating point for body water balance, they do not increase fluid retention or alter hydration markers (fluid balance) during exercise. However, the body of research evidence is limited, and more research is needed before firm conclusions can be drawn.

Sex based differences

When discussing differences between males and females, it is prudent to note differences in body mass, body composition, and, consequently, total body water. On average, males are heavier, with a higher fat-free mass (FFM) and therefore a higher total body water than females, as FFM contains ∼74% water compared with ∼10% in adipose tissue^{(Reference Visser, Gallagher and Deurenberg34)}. As a result, an equivalent absolute sweat loss represents a larger proportion of total body water in females. Moreover, cross-sectional studies investigating sex differences in sweat rate are often confounded by males’ greater body mass and FFM, which elevates exercise intensity and metabolic heat production, leading to higher reported sweat rates. Sweat rates are a consequence of the number of sweat glands and the amount of sweat produced per gland. While sweat gland distribution is similar between sexes, females may use a higher proportion of their glands but excrete less sweat per gland. This could improve thermoregulation efficiency, as a greater fraction of sweat is likely to evaporate, providing a cooling effect, rather than dripping off the skin, which has no thermoregulatory effect^{(Reference Avellini, Kamon and Krajewski35,Reference Shapiro, Pandolf and Avellini36)} .

For a full review of the sex differences on sweat losses, performance and thermoregulatory strain, the reader is directed to Wickham et al.^{(Reference Wickham, McCarthy and Spriet37)}. The authors suggest that women may experience a more rapid rise in core temperature at the onset of exercise and that thermoregulatory and cardiovascular strain associated with dehydration occur at lower levels of dehydration than in males. This may reflect lower total body water and blood volume, along with a smaller proportion of body fluid distributed to the extracellular compartment in females compared with males, which would amplify the physiological consequences of fluid loss^{(Reference Wickham, McCarthy and Spriet37)}. However, they do also highlight the methodological challenges when comparing fluid balance between sexes, including difficulties in matching for sweat loss, relative dehydration, training status and exercise intensity. Given these confounders, and the influence of exercise intensity, heat acclimatisation, environmental conditions and training status on fluid losses, it remains premature to draw firm conclusions about true sex differences. At present, the American College of Sports Medicine (ACSM) guidelines to individualise fluid intake based on measured or estimated losses is prudent advice to follow^{(Reference Thomas, Erdman and Burke38)}.

Sweat electrolyte concentrations

Sweat electrolyte losses reflect both sweat rate and electrolyte concentrations, with higher sweat rates generally associated with higher electrolyte concentrations and greater absolute losses. Consequently, given the typically lower sweat rates in females, their sweat electrolyte losses tend to be lower^{(Reference Baker, Ungaro and Sopena39)}. Using data from 1944 sweat tests, Baker et al.^{(Reference Baker, De Chavez and Nuccio40)} modelled the effects of various factors influencing sweat sodium concentration and found only a small effect of sex, concluding that sweat sodium cannot be reliably predicted without individualised sweat testing.

Fluid intakes

Thirst is primarily regulated by plasma osmolality. Vokes et al.^{(Reference Vokes, Weiss and Schreiber41)} showed that in 5 females measured in both the follicular and luteal phase, the onset of thirst occurs at a slightly lower plasma osmolality (by ∼5 mosm.kg⁻¹) in the luteal phase. Yet interestingly, this does not appear to influence ad libitum fluid intake between menstrual phases^{(Reference Freemas, Goss and Ables33)}. There may, however, be some sex-related differences in pre-exercise hydration. A systematic review of 24 studies on soccer players also shows a higher proportion of males classed as hypohydrated (urine specific gravity, USG ≥ 1.020), 66% male vs 47% of female players^{(Reference Chapelle, Tassignon and Rommers42)}. In line with these findings, Vope et al.^{(Reference Volpe, Poule and Bland43)} reported that a higher proportion of male athletes (47%) had a pre-practice USG ≥ 1.020 (hypohydrated) than female athletes (28%). However, findings are not universal; a smaller study among lower-level players (National Collegiate Athletic Association (NCAA) Division II basketball players) reported no sex differences^{(Reference Thigpen, Green and O’Neal44)}. Although USG has been criticised as a marker of hydration, it still provides an indication of hydration status^{(Reference Baker, De Chavez and Nuccio40)}.

Sex-differences have been reported in the incidence of exercise-associated hyponatremia (EAH; blood sodium < 135 mmol/L), a rare condition mainly caused by excessive fluid intake^{(Reference Almond, Shin and Fortescue45)}. Almond et al.^{(Reference Almond, Shin and Fortescue45)} analysed blood samples collected at the finish line of the Boston Marathon from 488 of the 766 runners, 13% of whom had EAH. Univariate analyses showed EAH was associated with substantial weight gain, fluid intake >3 L during the race, consumption of fluids at every mile, a racing time of >4 h, female sex and low BMI. However, once body mass, BMI and race time were controlled for, the apparent sex difference in EAH incidence disappeared^{(Reference Almond, Shin and Fortescue45)}. It should be noted that the studies mentioned are observational, and there are no controlled intervention studies evaluating the risk of EAH. Further, there is currently no evidence that changes in estradiol and/or progesterone levels in fluid volume regulation contribute directly to EAH. In fact, the ability of oestrogen to cause cell swelling by inhibiting the sodium-potassium ATP pump has been demonstrated only in rodent models^{(Reference Fraser, Kucharczyk and Arieff46,Reference Fraser and Sarnacki47)} . Therefore, education and behavioural change strategies focused on individual fluid requirements may aid fluid balance of females during endurance events, particularly given the multiple factors that affect both sweat and electrolyte losses, which make hydration requirements highly individual.

Sex differences

Details of studies investigating sex difference in carbohydrate need are shown in Table 3 (see supplementary material). Higher absolute rates of total CHO and fat oxidation have been reported in males compared with females^{(Reference Blatchford, Knowlton and Schneider48–Reference Tarnopolsky, MacDougall and Atkinson54)} in fasted conditions or when a placebo has been provided. These results have been attributed to a higher ratio of type 2 glycolytic muscle fibres in males than in females^{(Reference Tarnopolsky55)}.

A greater reliance on CHO during exercise amongst males has been associated with differences in the relative muscle fibre type ratio between sexes. Tarnopolsky^{(Reference Tarnopolsky55)}, Steffensen, Roepstorff^{(Reference Steffensen, Roepstorff and Madsen56)} and Hoeg, Roepstorff^{(Reference Hoeg, Roepstorff and Thiele57)} all report a greater percentage of type I fibres in females than in males. Therefore, males have a greater percentage of type II glycolytic fibres; these fibres have a greater dependency on CHO to fuel glycolytic processes^{(Reference Steffensen, Roepstorff and Madsen56–Reference Riddell, Partington and Stupka58)}.

Riddell, Partington^{(Reference Riddell, Partington and Stupka58)} and Campbell, Angus^{(Reference Campbell, Angus and Febbraio59)} have shown that ingestion of CHO suppresses the contribution of endogenous CHO to total energy expenditure during exercise amongst females but not in males. This may be due to the metabolic effects of 17β-estradiol during exercise. The administration of 17β-estradiol to males has been shown to spare muscle and liver glycogen during exercise^{(Reference Kendrick and Ellis60,Reference Rooney, Kendrick and Carlson61)} . In other human studies, 17β-estradiol has been shown to decrease the rate of glucose appearance and disappearance during exercise in amenorrheic females^{(Reference Ruby, Robergs and Waters62)}. However, these studies may have been influenced by a lack of strict control in participant selection and matching. Other studies have reached conclusions on differences in substrate oxidation between the sexes without making a direct in study comparison of the two sexes^{(Reference Friedlander, Casazza and Horning50)}.

It is important to ensure studies comparing males and females have appropriately matched participants to account for confounding factors that could influence results, such as body composition, training status and fitness levels. For example, when expressed relative to lean body mass (LBM), the greater reliance on CHO in male participants is diminished compared with females when a placebo is provided, with trials showing similar rates of CHO oxidation in metabolically active tissues Wallis, Dawson^{(Reference Wallis, Dawson and Achten63)}. Wallis et al. reported similar rates of total CHO oxidation relative to LBM between males and females during 120 minutes of exercise when CHO was not ingested. This study matched participants on relative VO_2max and training history, ensuring similar relative exercise intensities and subsequently relative energy expenditure.

Others have reported no difference in fat oxidation between sexes^{(Reference Wallis, Dawson and Achten63,Reference Costill, Fink and Getchell64)} when conducted at a higher exercise intensity than those that have reported sex-based differences^{(Reference Blatchford, Knowlton and Schneider48–Reference Tarnopolsky, MacDougall and Atkinson54,Reference Tarnopolsky, Atkinson and Phillips65)} . Exercise intensity has previously been shown to influence substrate oxidation. Of the studies that have shown differences in fat and CHO oxidation between males and females, exercise intensities range from 40% to 65% VO_2max. At these lower exercise intensities, fat is favoured as an energy substrate in both males and females^{(Reference Achten, Gleeson and Jeukendrup66–Reference Venables, Achten and Jeukendrup71)}. In studies where no differences in substrate oxidation have been observed in a fasted state^{(Reference Wallis, Dawson and Achten63,Reference Costill, Fink and Getchell64)} , the exercise intensity was higher, between 67% and 70% VO_2max, an intensity that begins to favour CHO oxidation and may subsequently reduce the difference in fat oxidation between males and females. Hackney, McCracken-Compton^{(Reference Hackney, McCracken-Compton and Ainsworth72)} reported differences in substrate oxidation at lower exercise intensities up to 60% VO_2max, but not in higher exercise intensities.

Menstrual phases

Exercise intensity has an important regulatory effect on substrate metabolism during exercise; in addition, it appears to affect the actions of female steroid hormones on substrate metabolism. Differences in substrate oxidation between menstrual cycle phases observed by Hackney, McCracken-Compton^{(Reference Hackney, McCracken-Compton and Ainsworth72)} at 35% and 65% VO_2max were diminished at 75% VO_2max. No differences in substrate oxidation across the menstrual cycle during higher intensity exercise have also been reported by other authors^{(Reference Bailey, Zacher and Mittleman73–Reference McLay, Thomson and Williams77)}. By contrast, Campbell and Febbraio^{(Reference Campbell and Febbraio78)} and Zderic, Coggan^{(Reference Zderic, Coggan and Ruby79)} reported higher lipid and lower CHO oxidation rates at higher exercise intensities in the luteal phase, see Table 4 (see supplementary material). However, given the challenges of determining which phase of the menstrual cycle a woman is in, even if these differences existed, it would be difficult to implement any dietary changes. Furthermore, the impact of varying CHO intake throughout the menstrual cycle on performance and health has not been investigated. Indeed, a study investigating performance with CHO ingestion across the menstrual cycle reported no differences between the luteal and follicular phases. Bailey et al. investigated differences in time to exhaustion amongst nine healthy, moderately trained women. Participants were tested twice in the follicular and twice in the luteal phase, during which they cycled at 70% of peak O₂ consumption until exhaustion. During the trials, they consumed 0.6 g CHO kg BM⁻¹⋅h⁻¹ (5 ml/kg of a 6% CHO solution every 30 min, beginning at min 30 of exercise) or a placebo drink. Time to exhaustion during the CHO trial increased from placebo values (P < 0.05) by 14.4 ± 8.5 (Follicular) and 11.4 ± 7.1% (Luteal); however, no differences were observed between menstrual cycle phases^{(Reference Bailey, Zacher and Mittleman73)}.

Summary

There is limited research on the effects of CHO ingestion on performance across menstrual phases; however, a systematic review found that performance does not differ between phases, suggesting that CHO needs are unlikely to differ. Obtaining differences between the sexes for oxidation rates is difficult, given differences in training status, VO₂max and body composition, which are generally seen between the sexes and could confound research findings. Further differences in substrate utilisation between phases and between sexes seem to disappear as intensity increases. Given the challenges in determining menstrual phase and the positive effects of CHO ingestion during endurance exercise, it would seem prudent to encourage CHO intake throughout the menstrual cycle, with amounts dependent on the individual’s goals and tolerability.

Protein before exercise

There is no evidence that consuming protein before exercise has any beneficial effects on recovery or protein synthesis compared with post-exercise consumption^{(Reference Schoenfeld and Aragon80)}. Commonly cited studies investigate the effects of pre- and post-exercise protein consumption rather than pre-exercise alone, see Table 5 (see supplementary material). For example, a study of 66 overweight participants (50% female) assigned to one of four groups for 8 weeks: (1) HIIT, 2 days a week; (2) essential amino acid (EAA) supplementation, 3.6 g twice daily before and after exercise; (3) HIIT + EAA; or (4) control. Participants were measured for body composition, RMR, substrate metabolism (RER) and cardiorespiratory fitness (VO_2max) at baseline, 4 weeks and 8 weeks^{(Reference Hirsch, Greenwalt and Cabre81)}. The authors concluded that the addition of twice-daily EAA did not seem to significantly enhance metabolic adaptations to HIIT, nor was there a sex interaction^{(Reference Hirsch, Greenwalt and Cabre81)}. Another study investigating EAA again before and after exercise with thirty-seven untrained adults (51% female) during eight weeks of HIIT, twice a week on a cycle ergometer, one group were supplemented with an EAA supplementation (HIIT + EAA; 3.6 g of EAA twice daily, 30 minutes pre and post HIIT) or whereas the other group performed the exercise without supplementation (HIIT)^{(Reference Hirsch, Cabre and Gould82)}. Again, there were no differences between the groups (Time to Exhaustion or workload progression) at any time point (1, 4 and 8 weeks). Nor was there any difference in heart rate nor Rating of Perceived Exertion (RPE) for the whole group at any time point, but there was a significantly lower RPE in females with EAA supplementation at 4 weeks, which was no longer present at 8 weeks^{(Reference Hirsch, Cabre and Gould82)}. Another study cited as evidence for beneficial effects of pre-exercise protein investigated a pre-exercise supplement (caffeine/beta-hydroxy-beta-methylbutyrate (HMB)/vitamin D) and post-exercise supplement (15 g whey protein, 5 g casein protein 20g CHO, vitamins C&D and glucosamine) compared to a no calorie placebo and a control group amongst 64 untrained males (n = 23) and females (n = 41) during a 6 week HIIT intervention twice a week^{(Reference Cabre, Gordon and Patterson83)}. There was no significant effect between the supplemented group, placebo group nor control for any body composition measure nor VO_2max when examined overall, nor for males nor females. There was a significant difference in the increase in leg press 1 repetition maximum (RM) between the supplemented group and the control group for the total sample. For females, there was also a significant difference in the increase in 1RM between the placebo and control groups. For upper-body strength, chest press 1 RM increased significantly more in the placebo group than in the control group among females only. There were no significant differences in the change in countermovement jump between groups^{(Reference Cabre, Gordon and Patterson83)}. Combined, given the lack of significant findings between placebo and interventions, these studies do not support a beneficial effect of pre-exercise protein compared to placebo. Although research on the impact of pre-exercise protein intake in women is limited, the ACSM importantly states, ‘Choices high in fat/protein/fibre may need to be avoided to reduce risk of gastrointestinal issues during the event’^{(Reference Thomas, Erdman and Burke38)}, a key consideration for training and competition. Most protein researchers agree that total daily protein intake is the primary driver of muscle protein synthesis and repair, rather than timing around exercise^{(Reference Schoenfeld and Aragon80)}. Moreover, research investigating protein timing before or after training report no effect on body composition^{(Reference Casuso and Goossens84)}.

Protein post-exercise

Although much of the research on exercise-induced anabolic responses has been conducted in males, recent research suggests that the ‘anabolic window’ is much longer than previously thought^{(Reference Schoenfeld and Aragon80)}, with muscle protein synthesis believed to remain elevated for at least 24 hours following resistance exercise^{(Reference Schoenfeld and Aragon80)}. Importantly, no sex differences in muscle protein synthesis have been observed following resistance training^{(Reference West, Burd and Churchward-Venne85)}. Systematic reviews and meta-analyses similarly indicate that males and females have comparable hypertrophic and lower-body strength adaptations in response to resistance exercise, although untrained females may show greater potential for upper-body strength^{(Reference Roberts, Nuckols and Krieger86)}.

There is some, albeit limited, data in females that nutrient timing around exercise does not impact body composition nor strength. Pihoker et al.^{(Reference Pihoker, Peterjohn and Trexler12)} investigated the effects of consuming a CHO-protein supplement either before or after resistance training versus no supplement for 5 weeks (10 HIRT sessions) amongst 43 trained females. They found no significant differences in body composition (measured via dual-energy X-ray absorptiometry, DEXA) or leg press strength (1RM). Whilst bench press was significantly greater in the pre- and post-trials compared to no supplement, the authors believed this was due to the novelty of the exercise in the participants.

Summary

While we acknowledge that research on nutrient timing in females is limited, with much of the cited research being mixed-sex rather than female-specific, the available evidence indicates that hormonal changes depend on the study protocol employed. Resistance training adaptations are broadly similar between males and females, and nutrient timing does not appear to affect body composition or lower-body strength. Some effects on upper-body strength have been observed, though further research is needed to substantiate them. Overall, current evidence suggests that achieving adequate total daily protein intake is the primary driver for muscle protein synthesis, with nutrient timing offering only limited additional benefit. However, for athletes training multiple times per day, consuming protein soon after exercise can help initiate recovery in preparation for the next training session.

Limitations of research to date

Although it is often stated in the media that there is no research in women, this narrative review shows that there is some, albeit limited, research in women. It should also be noted that much of the research has not quantified menstrual status or confirmed the phase of the menstrual cycle via hormonal testing, and none has followed the recommended gold standard for researching females^{(Reference Elliott-Sale, Minahan and de Jonge32)}. Further, the activity levels nor fitness levels of participants are not always described, or in some cases inactive individuals undertake a training programme, it is possible that the response of those unfamiliar with exercise may vary from those experienced exercisers.