Physiological sex differences affect the integrative response to exercise: acute and chronic implications

What is the topic of this review? We review sex differences within physiological systems implicated in exercise performance; specifically, how they integrate to determine metabolic thresholds and fatigability. Thereafter, we discuss the implications that these sex differences might have for long‐term adaptation to exercise. What advances does it highlight? The review collates evidence from recent physiological studies that have investigated sex as a biological variable, demonstrating that the physiological response to equivalent ‘dosages’ of exercise is not the same in males and females; thus, highlighting the need to research diversity in physiological responses to interventions.

improved, the body of empirical evidence has grown, and the previous belief that women are less resilient to physical fatigue conflicts with present-day understanding (Hunter, 2014). Despite the progress, the number of male-only studies in biological research vastly outnumbers female-only studies, particularly in physiology (∼4:1); moreover, when studies have included both sexes, only one-third have factored sex into their analyses (Beery & Zucker, 2011). By highlighting evidence demonstrating that the physiological responses to exercise differ between males and females, we will outline why the inclusion of sex as a biological variable is important to consider for optimal prescription of exercise in health and disease.
When an acute bout of exercise is considered, researchers often investigate how well an individual or a population can tolerate a given intensity, in addition to the physiological systems and/or perceptions that limit performance (Enoka & Duchateau, 2016;Hunter, 2018;Hureau, Romer, & Amann, 2018;. Acute exercise imposes a stress upon physiological systems that can elicit a cascade of molecular responses, which, when repeated and progressed over the course of a training programme, can lead to physiological adaptation and improved capacity for exercise (Granata, Jamnick, & Bishop, 2018). However, much like all interventions that target health improvements, the 'dosage' of exercise is important in determining training outcomes. Exercise modality, intensity, duration and frequency are all considered to mediate the adaptation elicited by a training programme (Bishop, Granata, & Eynon, 2014), in addition to the characteristics of the exercise performer(s) (Bouchard & Rankinen, 2001). The aim of this review is to highlight how biological sex influences properties of the physiological systems involved in exercise performance and how sex differences mediate the response to acute exercise. Furthermore, we detail how sex could affect adaptations to chronic exercise, with the implications for athletic performance, health and disease of males and females.

Skeletal muscle
The contractile apparatus of the motor pathway is ultimately responsible for force generation and human movement (Dulhunty, 2006). Reductions in the ability of skeletal muscle to produce force are frequently observed after various forms of locomotor exercise (Brownstein, Millet, & Thomas, 2020) and are considered to limit the ability of an individual to sustain a given intensity of exercise, particularly when the intensity is high (Burnley & Jones, 2018). Although the scope of the present review is not to detail all components of exercise-induced impairment to skeletal muscle contraction, we direct the reader towards previous seminal reviews of this topic (Allen, Lamb, & Westerblad, 2008;Fitts, 2008;Powers & Jackson, 2008). The characteristics of the skeletal muscle performing exercise are thought to influence energy provision, resistance to fatigue and long-term adaptability to exercise (Westerblad, Bruton,

New Findings
• What is the topic of this review?
We review sex differences within physiological systems implicated in exercise performance; specifically, how they integrate to determine metabolic thresholds and fatigability. Thereafter, we discuss the implications that these sex differences might have for long-term adaptation to exercise.
• What advances does it highlight?
The review collates evidence from recent physiological studies that have investigated sex as a biological variable, demonstrating that the physiological response to equivalent 'dosages' of exercise is not the same in males and females; thus, highlighting the need to research diversity in physiological responses to interventions.
& Katz, 2010); therefore, any sex difference in these physiological qualities could influence responses to acute and chronic exercise.
One simple observation is that males possess a greater quantity of skeletal muscle than females, which contributes to greater maximal strength (Ivey et al., 2000;Lindle et al., 1997;Miller, MacDougall, Tarnopolsky, & Sale, 1993;Welle, Tawil, & Thornton, 2008). When the physiological properties of skeletal muscle are investigated, sex differences that might influence the response to exercise become apparent. Sexual dimorphisms of gene expression lead to differences in phenotypic expression, with considerable differences in the morphological composition of skeletal muscle (Welle et al., 2008).
For instance, female vastus lateralis muscles are consistently shown to have a 7-23% greater proportional area of type I muscle fibres (Roepstorff et al., 2006;Simoneau & Bouchard, 1989;Staron et al., 2000). The consequences of this greater proportion of type I fibres are multifactorial; in terms of muscle metabolism during exercise, females oxidize more fat but less carbohydrate and amino acids compared with males (Tarnopolsky, 2008) and demonstrate faster oxygen uptake (V O 2 ) kinetics during moderate exercise intensities (Beltrame, Villar, & Hughson, 2017). This is probably attributable to the fact that males have greater glycolytic capacity (Esbjörnsson, Sylvén, Holm, & Jansson, 1993), whereas females have greater wholemuscle oxidative capacity (Russ, Lanza, Rothman, & Kent-Braun, 2005). Likewise, previous observations suggest that females exhibit a smaller decrease in ATP concentrations and increases in products of ATP breakdown after all-out exercise (Esbjornsson-Liljedahl, Bodin, & Jansson, 2002;Esbjörnsson-Liljedahl, Norman, Jansson, & Sundberg, 1999). Furthermore, trained males and females differ in mitochondrial oxidative function, with females demonstrating approximately onethird greater mitochondrial intrinsic respiratory rates than males as assessed via high-resolution respirometry after muscle biopsy (Cardinale et al., 2018). Collectively, these findings imply that female skeletal muscle metabolism is more suited to resynthesizing ATP from oxidative phosphorylation during exercise, and this is explored in more detail below.
The proportional fibre-type difference between sexes also influences the contractile properties of skeletal muscle in males and females. Male skeletal muscle exhibits faster relaxation rates compared with female muscle (Hunter, Butler, Todd, Gandevia, & Taylor, 2006), consistent with a greater proportional area of type I fibres in females. Type I fibres also demonstrate slower Ca 2+ kinetics, lower power generation and slower shortening and relaxation velocities than type II fibres (Schiaffino & Reggiani, 2011). Indeed, evidence from muscle biopsies has shown that females have a lower rate of maximal sarcoplasmic reticulum Ca 2+ -ATPase activity than males (Harmer et al., 2014). This difference in ionic regulation of muscle contraction has been suggested to contribute to a greater fatigue resistance of female muscle during exercise (Hunter, 2014(Hunter, , 2016a and is explored in more detail later in this review. Lastly, haemodynamic and perfusive properties of the muscle differ between sexes, possibly affecting oxygen delivery during exercise (Beltrame et al., 2017). Vasodilatory responses of the feed arteries to exercising skeletal muscles are greater in females; for instance, the femoral artery exhibits greater vascular conductance and blood flow during incremental exercise (Parker et al., 2007). This sex difference could promote greater muscle perfusion during exercise; however, this would also depend on the capillarization of muscle. Evidence from muscle biopsies demonstrates a higher density of capillaries per unit of skeletal muscle in females compared with males in the vastus lateralis (Roepstorff et al., 2006), a key factor of aerobic exercise performance (Mitchell, Martin, Bailey, & Ferguson, 2018a;Vanhatalo et al., 2016).
Collectively, evidence suggests that the metabolic, contractile and haemodynamic properties of skeletal muscle differ between males and females, which mediates the resultant response to the same 'dosage' of exercise.

Respiratory system
In healthy males and females, the respiratory system is relatively well equipped to tolerate high intensities of exercise. Key muscles within the respiratory system, such as the diaphragm, are purportedly the most fatigue resistant of all skeletal muscles, with a high aerobic enzymatic capacity, blood supply and resistance to vasoconstriction (Dempsey, Romer, Rodman, Miller, & Smith, 2006). Furthermore, the increased recruitment of accessory inspiratory and expiratory muscles with increasing exercise intensities alleviates the workload on individual muscles, such as the diaphragm (Aliverti et al., 1997).
Despite these fatigue-resistant morphologies, respiratory muscle fatigue occurs during exercise above ∼80% of maximal ramp test power (Johnson, Babcock, Suman, & Dempsey, 1993), which can affect the integrative response of physiological systems implicated in exercise performance (Romer & Polkey, 2008;Sheel, Boushel, & Dempsey, 2018) and, possibly, in a different manner in males and females. Specifically, as all exercising skeletal muscles 'compete' for available cardiac output, the respiratory muscle metaboreflex initiates sympathetic vasoconstrictor outflow to exercising skeletal muscle, reducing limb blood flow (Sheel et al., 2001), ultimately exacerbating fatigue of locomotor muscles and limiting whole-body exercise performance (Harms, Wetter, Croix, Pegelow, & Dempsey, 2000;Mador & Acevedo, 1991 (Guenette, Witt, McKenzie, Road, & Sheel, 2007). This is considered to be a result of the smaller airway size, as the phenomenon only becomes apparent at elevated V˙E .
If theV O 2 of a muscle is directly proportional to blood flow in the Fick equation, respiratory musculature is likely to require a greater fraction of cardiac output for females . Investigation into sex differences in blood flow competition is limited, although greater demand for blood at the respiratory level is likely to have detrimental effects on locomotor muscle blood flow (Romer & Polkey, 2008;Sheel et al., 2018).
Exercise-induced arterial hypoxaemia can occur at high exercise intensities, as a result of many factors, including alveolar hypoventilation, diffusion limitation, ventilation-perfusion mismatch and cardiac and pulmonary shunt (Dempsey & Wagner, 1999).
It was suggested that females might be more prone to exerciseinduced arterial hypoxaemia than males owing to their smaller lung volumes, smaller airway sizes and fewer alveoli Harms et al., 1998). When matched for lung size, no sex difference was observed in pulmonary diffusive capacity (Bouwsema, Tedjasaputra, & Stickland, 2017) and ventilation-perfusion mismatch (Olfert et al., 2004). However, at a population level, the observed difference in lung size between males and females could contribute to the greater prevalence of exercise-induced arterial hypoxaemia in females.
Despite these potentially more limiting morphologies within the female respiratory system compared with males, some functional aspects could counteract any central limitation to exercise performance. For example, the diaphragm is more fatigue resistant in females compared with males (Guenette et al., 2010;Welch, Archiza, Guenette, West, & Sheel, 2018). Furthermore, females appear to recruit extra-diaphragmatic accessory muscles at elevated V˙E, which could also serve to alleviate diaphragm fatigue (Mitchell, Schaeffer, Ramsook, Wilkie, & Guenette, 2018b). The greater fatigue resistance of the diaphragm appears to elicit a lower respiratory muscle metaboreflex in females compared with males, preserving lower limb blood flow and vascular resistance (Smith et al., 2016(Smith et al., , 2017. The caveat exists that the findings of Smith et al. (2016Smith et al. ( , 2017 were after resistive breathing; therefore, whether a similar sex difference exists during whole-body exercise is currently unknown. Despite this greater fatigue resistance, the evidence regarding sex differences within the respiratory system could lead to the suggestion that central limitations to whole-body exercise performance might be greater in females. The implications of these sex differences for acute and chronic exercise prescription are discussed below in Section 4.

Nervous system
Muscle contraction is initiated by the CNS, with muscle activation being a limiting factor to exercise performance in selected situations.
Long durations of exercise are typically associated with CNS impairment (Martin et al., 2010;Thomas et al., 2014Thomas et al., , 2016, concerning mechanisms from a variety of sites (Gandevia, 2001).
Changes in neurotransmitter concentrations within the motor cortex (Meeusen & Roelands, 2018) and motoneurons (Kavanagh, McFarland, & Taylor, 2019) influence the ability to drive the muscle. Excitability of descending tracts can enhance or reduce the amount of neural drive for a given synaptic input (Sidhu et al., 2017(Sidhu et al., , 2018Weavil, Sidhu, Mangum, Richardson, & Amann, 2016), and the firing rates of motor units can decline (Woods, Furbush, & Bigland-Ritchie, 1987). The combination of these physiological occurrences contributes to a reduction in voluntary activation (VA; Enoka, 2012), commonly measured during maximal contractions using the twitch-interpolation technique and termed central fatigue (Merton, 1954). In terms of sex comparisons of changes in VA in response to exercise, the evidence is conflicting, which is, perhaps, a reflection of the variety of tasks used to investigate potential sex differences. After fatiguing tasks involving maximal isometric contractions, a greater reduction in VA was observed in males compared with females (Martin & Rattey, 2007;Russ & Kent-Braun, 2003;Senefeld, Pereira, Elliott, Yoon, & Hunter, 2018). Although tasks involving submaximal contractions have not demonstrated this difference in pre-versus postexercise values, VA showed a slower rate of decrease in females than in males during the task (Ansdell et al., 2019a). Mechanistically, this sex difference could be a combination of various neurophysiological occurrences; however, the fact that it occurs during maximal contractions could align with the hypothesis that females experience less blood flow occlusion during high-intensity contractions (Hicks, Kent-Braun, & Ditor, 2001;Hunter, Griffith, Schlachter, & Kufahl, 2009). A greater ability to maintain blood flow during maximal contractions would result in a lesser/slower accumulation of nociceptive metabolites during exercise and a subsequent reduction in activation of group III/IV afferents (Hunter, 2014). It has been suggested that activation of these neurons increases intracortical inhibition (Hilty et al., 2011;Sidhu et al., 2018) and reduces cortical and spinal excitability (Sidhu et al., 2017). These factors impair the synaptic input to the motoneuron in addition to affecting motoneuronal transmission of motor commands directly, resulting in an impaired ability to activate the muscle (Sidhu et al., 2017). Therefore, the sex difference in VA might not result from differences in intrinsic CNS properties but might instead be a reflection of slower metabolic disturbance within skeletal muscle.
Despite this, the evidence regarding the potential sex difference in CNS function during isometric contractions does little to inform our understanding of the response to whole-body exercise. As has been emphasized in the literature, in order to ascertain the neurophysiological response to an exercise task, the assessment task must be the same as the exercise task (Brownstein et al., 2018;Kalmar, 2018;Sidhu, Cresswell, & Carroll, 2013). Indeed, as discussed by Weavil and Amann (2018), corticospinal neurons are modulated by a myriad of extraneous factors, such as hydration status (Bowtell, Avenell, Hunter, & Mileva, 2013), ventilatory rate (Hopkinson et al., 2004) and arterial blood gas concentration (Hartley et al., 2016). Conceivably, any sex difference in these processes could influence the excitability and function of the CNS. Evidence has shown less of a decline in VA after cycling exercise below critical power (CP) compared with above CP, despite a lack of difference in the excitability of descending tracts between males and females . However, further investigation is required into this topic; likewise, the influence of the menstrual cycle on CNS function during exercise is currently unknown.
The motor unit is the final motor pathway of the nervous system and directly influences force production by the muscle (Del Vecchio et al., 2018. There are few studies, however, determining whether there are sex differences in motor unit behaviour. Higher Males have a lower mechanical work of breathing, and oxygen cost of breathing at high minute ventilations.
The limitations of the respiratory system potentially contribute to a greater risk of exercise-induced arterial hypoxaemia in females.

Respiratory System
Males have larger lungs and airways, even when height matched.

Supraspinal Neurons
Intracortical inhibition and voluntary activation influenced by hormonal status in females.

Circulatory System
Males have higher concentrations of haemoglobin than females.

Motoneurons
Females have greater proportion of low-threshold motor units, and greater oscillations in the common synaptic input to motor units.

Skeletal Muscle
Females typically have greater proportional area of type I muscle fibres. Therefore, greater capillarization, mitochondrial respiratory capacity and fatigue resistance.
F I G U R E 1 Summary of the sex differences within the key physiological systems implicated in exercise performance discharge rates were observed in males compared with females at low (15%) contraction intensities (Harwood, Cornett, Edwards, Brown, & Jakobi, 2014), possibly owing to a lower proportion of low-threshold motor units in males compared with females (Kukulka & Clamann, 1981). These factors, amongst others, are suggested to contribute to a sex difference in force steadiness (Jakobi, Haynes, & Smart, 2018), although the authors concluded that despite a clear sex difference in steadiness, the underlying mechanisms are unclear. One recent investigation using high-density surface electromyography (Pereira et al., 2019) demonstrated no differences in discharge rate, but showed that females experienced greater oscillations in the common synaptic input to motor units. This greater coefficient of variation in the common synaptic input has been associated with the coefficient of variation of force during sustained submaximal contractions (Farina & Negro, 2015). The present challenge is to assess how the interaction between sex, motor unit firing properties and force steadiness affect functional tasks, which is challenging (Jakobi et al., 2018).

SEX DIFFERENCES IN THE INTEGRATION OF PHYSIOLOGICAL SYSTEMS DURING EXERCISE
The integration of physiological systems is necessary for the regulation of homeostasis during fatiguing exercise. Performance fatigability, which is defined as a decline in an objective measure (e.g. maximal force/power production; Enoka & Duchateau, 2016;Hunter, 2018), in particular, will be determined by the cardiorespiratory and neuromuscular systems working to enhance the transport of oxygen from air to muscle for a given exercise intensity (Amann & Calbet, 2008).
Although there are sex differences, from the first stages of the oxygen cascade (i.e. ventilation and gas exchange) to the final stages of oxygen usage within the muscle (see Figure 1), there is limited evidence for potential sex differences in the integrative response of physiological systems during exercise. This section will outline how sex differences in the physiological systems implicated in exercise performance integrate to determine the metabolic response to exercise, in addition to the influence that this has on commonly used tests of aerobic and anaerobic performance.
Studies investigating sex differences in performance fatigability of whole-body exercise have typically used self-paced tasks (Glace, Kremenic, & McHugh, 2013;Temesi et al., 2015). This approach, however, can hinder insight into the rate-limiting mechanisms of muscle metabolism because of differing rates of fatigue throughout a pacing strategy (de Almeida Azevedo et al., 2019). An alternative approach is constant-load exercise, with exercise intensity normalized within intensity domains, such as the power-duration relationship (see subsection 3.1), because the metabolic demands of exercise will remain consistent within and between groups. The relationship between exercise intensity and duration presents an opportunity to study the physiological mechanisms underpinning exercise performance (Jones, Vanhatalo, Burnley, Morton, & Poole, 2010) and allows mechanistic comparison between exercise tasks and even between species. For this reason, our group recently investigated the sex difference in the integrative response to exercise by using the intensity-duration relationship during single-limb isometric (Ansdell et al., 2019a) and whole-body locomotor exercise (Ansdell et al., 2020).

The power-duration relationship
Many investigations into the power-duration relationship have been performed with male participants only, with exercise intensity typically categorized into four distinct 'domains' , each of which elicits different profiles of neuromuscular fatigue (Burnley & Jones, 2018).
Below CP (i.e. the heavy-intensity domain), metabolic responses reach a steady state, as described above, but above CP the muscle metabolic milieu increases progressively until task failure is reached (Vanhatalo, Fulford, Dimenna, & Jones, 2010). An increase in exercise intensity above CP increases the rate of muscle metabolic perturbation ; however, the end-exercise metabolic milieu and consequent contractile dysfunction are remarkably consistent, regardless of work rate Jones, Wilkerson, DiMenna, Fulford, & Poole, 2008;Schäfer, Hayes, & Dekerle, 2019;Vanhatalo et al., 2010Vanhatalo et al., , 2016. Exercise within the severe domain terminates soon after attainment of maximal oxygen uptake (V O 2 max ). Exercise below CP, in the heavy-intensity domain, results in metabolic perturbation within the exercising muscle; however, this reaches a submaximal steady state and is therefore unlikely to be a limiting factor to exercise Jones et al., 2008;Vanhatalo et al., 2010).
This border between intensity domains is one parameter within the power-duration relationship and, together with CP, the curvature constant (W′) allows insight into different physiological mechanisms.
The CP can be calculated from the hyperbolic relationship between exercise intensity (e.g. speed or power) and the time to task failure (Hill, 1925;Hill & Smith, 1999;Hill, Poole, & Smith, 2002;Jones et al., 2010;Monod & Scherrer, 1965); it is considered to be the greatest exercise intensity that results in 'wholly oxidative' energy provision (Poole, Burnley, Vanhatalo, Rossiter, & Jones, 2016). A crucial determinant of CP during whole-body exercise is delivery of oxygen to the skeletal muscle, and this is demonstrated by positive associations between CP and the fraction of inspired oxygen (Dekerle, Mucci, & Carter, 2012;Vanhatalo et al., 2010). Critical power is also positively correlated with the proportion of type I fibres and the muscle capillarity of knee-extensor muscles (Mitchell et al., 2018a;Vanhatalo et al., 2016).
Collectively, these studies indicate that the CP is dependent on the oxygen delivery to the muscles (convective factors) and oxygen extraction and utilization within the skeletal muscle (diffusive factors).
The mechanistic link between muscle morphology and CP has implications for understanding sex differences in exercise fatigability.
As previously explained, in comparison to males, females typically exhibit a greater proportional area of type I fibres and capillary density in the knee extensors (Roepstorff et al., 2006;Simoneau & Bouchard, 1989;Staron et al., 2000).  (Andersen, 1975). These factors combine to permit metabolic differences between fibres, such as a greater ability to regenerate ATP via the tricarboxylic acid (Krebs) cycle and therefore an ability to match ATP consumption with aerobic resynthesis in type I fibres, an occurrence that is not possible in type II fibres (Schiaffino & Reggiani, 2011  diffusive factors, such as muscle capillarity, mitochondrial density and the arteriovenous oxygen difference (Wagner, 1988(Wagner, , 1996. When a large muscle mass is engaged in the task (e.g. cycling or running), blood flow is distributed to different muscles and physiological systems during exercise by sympathetically mediated vasoconstriction in an attempt to maintain adequate perfusion pressures , and the required skeletal muscle blood flow can exceed the pumping capacity (cardiac output) of the heart (Calbet et al., 2004). As a result, the physiological challenges during whole-body exercise are multifactorial when compared with single-limb exercise, where the quantity and magnitude of muscles recruited are far less.
The CP during whole-body exercise, however, is dependent on both oxygen delivery and utilization Monod & Scherrer, 1965). This is demonstrated by experimental work showing that complete blood flow occlusion reduces abolishes CP (Broxterman et al., 2015), whereas manipulating the degree of occlusion by changing the contraction duty cycle shows a positive association between rest period duration (non-occlusion) and CP (Broxterman et al., 2014). Additionally, individuals with impaired oxygen delivery (e.g. people with central limitations, such as chronic obstructive pulmonary disease) exhibit a severely reduced CP (Van Der Vaart et al., 2014).
When this information is placed in the context of sex differences, it is important to consider the fact that females have smaller lung volumes (Schwartz et al., 1988) and lower haemoglobin concentrations (Cureton et al., 1986) compared with males. Together, these data indicate greater central limitations in females during high-intensity whole-body exercise (Harms et al., 1998), reducing oxygen delivery to the working muscle(s). Thus, the beneficial metabolic properties of female skeletal muscle (diffusive capacity) might counteract a reduced ability of females to deliver oxygen (convective capacity) relative to males during fatiguing whole-body exercise. Consequently, CP is shown to be similar between the sexes, even when expressed relative to a maximum (e.g. maximal ramp test power, P max ; Ansdell et al., 2020).
Furthermore, recent evidence for greater mitochondrial respiration in females than in males (Cardinale et al., 2018) suggests that this is an adaptation in the oxygen cascade to compensate for a lower oxygen delivery during exercise. Accordingly, the power-duration relationship during cycling is equivalent between the sexes.

Physiological responses to equivalent 'dosages' of exercise
Typically, when the responses of males and females are compared during constant-load exercise, the intensity is normalized to an arbitrary percentage of maximal capacity. In a single-limb model, exercise at a percentage of the maximal voluntary contraction (Ansdell, Thomas, Howatson, Hunter, & Goodall, 2018;Hunter et al., 2006Hunter et al., , 2009) is commonly used, whereas for locomotor exercise, percentages of maximal ramp test capacity (power for cycling, P max ) are used (Dominelli et al., 2017;Guenette et al., 2010). Although constantload exercise enables conclusions to be drawn about exercise work rate, normalization to arbitrary percentages of a maximum neglects the influence of metabolic thresholds. Consequently, in these studies, males and females could exercise at different work rates within an intensity domain or even in different exercise intensity domains. For example, during single-limb isometric exercise, if both sexes exercise at 50% maximal voluntary contraction (e.g. Ansdell et al., 2018), females, on average, are exercising at ∼118% of critical torque and males at ∼143% of critical torque (Ansdell et al., 2019a(Ansdell et al., , 2019b. As is well established, exercising at increasing intensities within the severe domain leads to a faster rate of metabolic disturbance . Therefore, the intensity of exercise is not matched between the sexes, and task failure occurs sooner in males than in females. During cycling exercise, where CP is not different between males and females , this difference is more subtle. Within a group of well-trained participants, CP typically occurs between 70 and 80% of P max (Poole, Ward, Gardner, & Whipp, 1988, 2016. If participants are subsequently studied at an intensity of 80% P max , for example, this could lead to participants working at different intensities within the severe domain or within distinct intensity domains, which elicit disparate metabolic responses (Jones comparisons. Discussion around the pros and cons of prescribing exercise intensity relative to metabolic thresholds (Mann, Lamberts, & Lambert, 2013), and the 'optimal' metabolic threshold (Jones et al., 2019), is available elsewhere. However, to compare the performance of metabolically matched exercise intensities between the sexes, threshold normalization is probably more appropriate.
In two recent experiments that took the threshold normalization approach, one single-limb (Ansdell et al., 2019a) and one cycling exercise , a consistent finding was that for the same duration and intensity of exercise, females experienced approximately half the reduction in contractile function of males in the knee extensors at the end of the fatiguing exercise (e.g. Figure 3b). During single-limb exercise normalized to the critical torque, this enabled females to exercise for approximately twice the duration before task failure (Ansdell et al., 2019a), whereas during cycling exercise, the time to task failure was similar between sexes .
The discrepancy between contractile impairment and the time to task failure in different exercise modalities is likely to be a result of different limiting factors of the two tasks. Severe-intensity exercise is thought to terminate once W′ is fully utilized and a 'critical' degree of metabolic disturbance within the exercising musculature is attained (Burnley & Jones, 2018;Jones et al., 2008;Vanhatalo et al., 2010), an interpretation that explains single-limb exercise tolerance (Broxterman et al., 2015). However, for whole-body exercise, the mechanisms are less clear. In terms of the power-duration relationship, severe-intensity exercise tolerance can be explained by W′; however, as stated by Poole et al. (2016), a physiological equivalent of W′ is unknown, and the nature of W′ cannot be limited to a single physiological process. It is possible that the depletion of W′ reflects a balance of cardiopulmonary and neuromuscular 'stress' , reflecting the sensory tolerance limit of afferent information from many physiological systems (Hureau et al., 2018;Thomas et al., 2018). As suggested by Poole et al. (2016), the constituent parts of W′ will differ between exercise tasks, in addition to the characteristics of the participants, which is relevant for the consideration of sex as a biological variable.
Although relative W′ is similar between males and females (and therefore severe-intensity time to task failure) during cycling, the reduced contractile dysfunction in females  suggests that the muscular component of W′ is smaller in females than in males. Conversely, based on the elevated W b and concurrent sensations of dyspnoea (Cory et al., 2015;Dominelli et al., 2015;Schaeffer et al., 2014), the respiratory component of W′ could be greater for females. This idea is demonstrated in Figure 4. In the subsequent sections, the implications of this potential sex difference are explained for acute and chronic exercise.

THE ACUTE AND CHRONIC IMPLICATIONS OF SEX DIFFERENCES
Optimizing the function of relevant physiological systems over the course of a training cycle is key to enhancing exercise performance, but this process relies upon the delivery of consistent, optimal, adaptive stimuli. This section will therefore discuss the mechanisms of aerobic adaptation after high-intensity exercise and how the sex differences in physiological responses during fatiguing exercise could influence this process.
High-intensity exercise induces a state of metabolic stress, the result of non-oxidative ATP resynthesis, which consists of an accumulation of metabolites and depletion of high-energy phosphate stores (Allen et al., 2008). This metabolic stress is thought to elicit long-term adaptations in the respiratory capacity of mitochondria via an increase in the electron transport system respiratory state (Jacobs et al., 2013;Lundby & Jacobs, 2016). The metabolites and the associated intracellular hypoxia provoked during highintensity exercise are potent stimuli for activation of AMP-activated A Females

B Males
Tolerable symptom of fatigue F I G U R E 4 A simplified conceptual model to demonstrate that although the maximal tolerable degree of fatigue is similar between males and females, the magnitudes of its constituent parts are different. Adapted from Thomas et al. (2018) protein kinase (AMPK) and hypoxia-inducible factor α (HIF-α) pathways. Likewise, the production of reactive oxidative species is thought to stimulate peroxisome proliferator-activated receptor gamma coactivator (PGC-1α) and p38 mitogen-activated protein kinase (p38MAPK) pathways (Ji, 2007). Collectively, these signalling responses are thought to be crucial stages in mitochondrial biogenesis and angiogenesis (Marcinko & Steinberg, 2014) and to be responsible for the expression of genes involved in troponin I and myoglobin content in mice (Lin et al., 2002) and humans (Olesen, Kiilerich, & Pilegaard, 2010).
Exercise within the severe-intensity domain is associated with what is described above as metabolic stress, and disruption to the metabolic environment within the muscle could be inferred from postexercise reductions in contractile function. Consistently in the studies by Ansdell et al. (2019aAnsdell et al. ( , 2020, it was demonstrated that for an equivalent duration of high-intensity exercise, females experienced less contractile dysfunction than males, potentially indicative of less disruption to the muscle metabolic environment and, perhaps, less of an adaptive stimulus for an equivalent amount of work. Evidence from proton magnetic resonance spectroscopic techniques during isometric exercise modalities would support this notion, indicating greater acidosis (Russ et al., 2005); however, equivalent data during locomotor exercise are difficult to obtain and, as yet, it is unknown whether the same phenomenon occurs.
There is a dearth of studies directly investigating the influence of sex on the adaptive signalling response after acute high-intensity exercise similar to a 'real-world' bout, such as interval training (e.g. multiple severe-intensity work periods, interspersed with moderateintensity recovery periods or rest). However, one study comparing the acute postexercise signalling response between males and females has previously used an 'all-out' Wingate test to elicit metabolic stress, with no difference in phosphorylation of a variety of signalling molecules (Fuentes et al., 2012). It is important to note that the postexercise response to other forms of aerobic exercise in females is mediated by the hormonal status of females, with a strong oestrogenic effect (Fu, Maher, Hamadeh, Ye, & Tarnopolsky, 2009). Such an effect might have confounded the findings of Fuentes et al. (2012), because they did not control for menstrual cycle phase and they mixed eumenorrhoeic females with oral contraceptive users, two practices that have recently been deemed 'inappropriate' (Sims & Heather, 2018) Further considerations for acute exercise prescription relate to the greater rate of recovery of contractile function postexercise in females compared with males (Ansdell et al., 2019a;Senefeld et al., 2018). Forms of interval exercise relying on repeated activity close to, or exceeding, the intensity elicitingV O 2 max are based upon the notion that repeated disruption of the metabolic environment of the exercising muscle(s) will augment aerobic and anaerobic adaptation (Gibala, 2009). It is possible that females require less rest between intense interval exertions for adequate metabolic recovery; if standard training recommendations that are based predominately on male research are applied to both sexes, females might not receive an optimal stimulus for adaptation. Indeed, a recent study investigated whether manipulation of the rest period duration affects the adaptation to an interval training programme in males and females (Schmitz et al., 2020). The authors demonstrated a reduced training effect in female participants when rest periods were extended from 30 to 180 s, with no such effect in males. Therefore, manipulating the configuration of an acute bout of exercise to the characteristics of the exercise performer, whether in a clinical or an athletic setting, is necessary to elicit the maximal adaptive response. Discussion concerning whether females and males require different training stimuli to augment aerobic performance is minimal: in their recent 'Connections' article in Experimental Physiology, Diaz-Canestro and Montero (2020) draw an analogy between the greater oxidative capacity of females and a 'superior combustion engine' , and they briefly discuss the implications of these differences within the oxygen cascade.
Future research could investigate the acute responses of the molecular adaptive pathways and further interrogate the plasticity of skeletal muscle in both sexes in response to acute and chronic highintensity exercise protocols. Tailoring the configuration of acute bouts of exercise to the characteristics of the performer is vital for achieving optimal adaptation to chronic exercise. One clinical example of where this approach has succeeded is with chronic obstructive pulmonary disease patients, who experience similar training adaptations with lesser exercise-related symptoms for interval exercise compared with continuous training (Vogiatzis, Nanas, & Roussos, 2002.
Although the vast majority of evidence presented in this review is based on healthy participants, the potential for sex to mediate exercise responses in clinical populations exists; however, the interaction between disease, sex and exercise is relatively under-explored and is likely to be different between diseases. Sex differences in physiology and responses to exercise are becoming increasingly evident, and the danger is that training programmes developed and optimized using only male data might not elicit equivalent adaptation in female populations.

THE INFLUENCE OF THE MENSTRUAL CYCLE
The effects of the menstrual cycle and the hormonal fluctuations both within a cycle and throughout the female reproductive lifespan (menarche to menopause; Brown & Thomas, 2011) are not well known. Furthermore, the use of exogenous hormones to control endogenous fluctuations, such as contraceptives/therapies, has the potential to affect any physiological system containing hormone receptors, including the CNS.
Within the CNS, sex hormones can act as 'neurosteroids' and modulate neuronal function (Paul & Purdy, 1992). Where exercise is concerned, this manifests as an altered ability to activate skeletal muscle (Ansdell et al., 2019d); however, the evidence concerning maximal strength and exercise performance fatigability is equivocal (Blagrove, Bruinvels, & Pedlar, 2020;Elliott-Sale et al., 2020;McNulty et al., 2020). In terms of skeletal muscle, and independent of exercise performance, contractility (Collins et al., 2018) and fatigability (Cabelka et al., 2019) appear to be influenced by oestrogen in mouse models, although substrate metabolism during exercise is thought to be mediated by menstrual cycle phase (Oosthuyse & Bosch, 2010). Within the respiratory system, V˙E and the ventilatory equivalent of carbon dioxide (V E ∕V CO 2 ) has been suggested to be increased within the luteal phase (Jurkowski, Jones, Toews, & Sutton, 1981;Schoene, Robertson, & Pierson, 1981), although this is not a consistent finding (Bemben, Salm, & Salm, 1995). Additionally, fluid regulation is altered by female sex hormone concentrations (Stachenfeld, 2008). These fluctuations within physiological systems could lead to the hypothesis that the integrative response to exercise might be altered by hormonal status, a hypothesis tested by Mattu, Iannetta, MacInnis, Doyle-Baker, and Murias (2020). The findings demonstrated no change in submaximal (V O 2 kinetics and the maximal lactate steady state) or maximal (V O 2 max and time to exhaustion test) responses to exercise across the menstrual cycle or between eumenorrhoeic females and combined oral contraceptive users (Mattu et al., 2020). It is important to note, however, that although metabolic thresholds and/or performance outcomes do not differ across the menstrual cycle, this does not mean that the integrative response to exercise does not change. As was demonstrated by Ansdell et al. (2020), when CP and time to exhaustion were not different between sexes, the response of physiological systems it is important to remember that when the integrative response to exercise across the menstrual cycle is considered, absence of evidence is not evidence of absence, and careful experimental design in future will enable the effects of changing sex hormone concentrations to be discerned.

CONCLUDING REMARKS
The consideration of sex in the prescription of exercise for males and females is key to optimizing adaptation in both athletic and clinical settings. Sex differences within the phenotype of physiological systems lead to differing responses to a host of interventions, and exercise is no exception. The ability to transport and utilize oxygen and differences in the fatigue resistance of contractile elements appear to contribute to the sex difference in the integrative response to an acute bout of exercise. The consequences of these sex differences in a long-term exercise programme are relatively under-researched in comparison; the possibility exists that training programmes based on data obtained from male participants leads to females receiving a potentially suboptimal exercise 'dosage' . Diversity exists across the physiological responses to a given intervention, and sex is simply one of the influencing factors in the puzzle to optimize outcomes for all.

Future investigations into how the biological context of an individual
influences acute and chronic physiological responses to exercise will enable optimal training patterns to be identified, promoting athletic excellence and health outcomes in all humans.

COMPETING INTERESTS
None declared.