Sex and age interaction in fundamental circulatory volumetric variables at peak working capacity

Whether the fundamental hematological and cardiac variables determining cardiorespiratory fitness and their intrinsic relationships are modulated by major constitutional factors, such as sex and age remains unresolved. Transthoracic echocardiography, central hemodynamics and pulmonary oxygen (O2) uptake were assessed in controlled conditions during submaximal and peak exercise (cycle ergometry) in 85 healthy young (20–44 year) and older (50–77) women and men matched by age-status and moderate-to-vigorous physical activity (MVPA) levels. Main outcomes such as peak left ventricular end-diastolic volume (LVEDVpeak), stroke volume (SVpeak), cardiac output (Qpeak) and O2 uptake (VO2peak), as well as blood volume (BV), BV–LVEDVpeak and LVEDVpeak–SVpeak relationships were determined with established methods. All individuals were non-smokers and non-obese, and MVPA levels were similar between sex and age groups (P ≥ 0.140). BV per kg of body weight did not differ between sexes (P ≥ 0.118), but was reduced with older age in men (P = 0.018). Key cardiac parameters normalized by body size (LVEDVpeak, SVpeak, Qpeak) were decreased in women compared with men irrespective of age (P ≤ 0.046). Older age per se curtailed Qpeak (P ≤ 0.022) due to lower heart rate (P < 0.001). In parallel, VO2peak was reduced with older age in both sexes (P < 0.001). The analysis of fundamental circulatory relationships revealed that older women require a higher BV for a given LVEDVpeak than older men (P = 0.024). Sex and age interact on the crucial circulatory relationship between total circulating BV and peak cardiac filling, with older women necessitating more BV to fill the exercising heart than age- and physical activity-matched men.


Background
Notwithstanding the undeniable complexity of the human body, the vital organ par excellence entails a remarkable mechanistic simplicity in its most fundamental function [1,2]. The direct relationship between cardiac filling and stroke volume (SV), known as the 'law of the heart' or Frank-Starling mechanism, was originally formulated at the dawn of twentieth century [3]. In essence, the more the heart is filled, the greater the output (Q = SV × heart rate), until a plateau is reached [1,2]. In humans, experimental evidence denotes that cardiac filling in the upright position is far beneath the plateau section, such that any change in left ventricular end-diastolic volume (LVEDV) alters SV proportionally [4]. In the absence of major cardiovascular abnormalities, cardiac filling primarily depends on the filling level of the circulatory system, i.e., blood volume (BV) [5]. BV modulates cardiac function at rest as well as at peak exercise, in which peak SV (SV peak ) exclusively determines cardiac output (Q peak ), since peak heart rate (HR peak ) is essentially uniform for a given age [6]. Importantly, outstanding levels of cardiorespiratory fitness, as represented by Q peak and its pulmonary correlate peak oxygen uptake (VO 2peak ), are strongly associated with large reductions in cardiovascular and overall mortality [7,8]. Invariably, high levels of cardiorespiratory fitness are accompanied by BV expansion [9,10]. Conversely, blood withdrawal induces proportional decrements in Q peak and VO 2peak , which do not recover until intravasscular volumes are restored [10][11][12], firmly supporting the crucial role of BV in the human circulatory system. The essential physiology of the circulatory system appears to be fairly straightforward. Yet, the fact that our current understanding is based on animal experiments and human studies mainly comprising healthy young men cannot be overlooked [13][14][15]. Moreover, the remarkable cardiovascular phenotypic adaptability to abnormal hemodynamic conditions, lifestyle stimuli as well as constitutional factors should be considered [16][17][18][19]. Most notably, whether the constitutional factors per excellence, i.e., sex and age, alter the fundamental circulatory relationships remains unresolved [16,19,20]. As a case in point, the inexorable age-dependent stiffening and remodeling of the cardiovascular system may interact with sex-specific hemodynamic and hematological characteristics [16,17,[21][22][23], plausibly altering the regulation of and relationship between primary volumetric variables in the systemic circulation [18,24,25]. While the prevailing emphasis on molecular cardiology is an ineluctable consequence of technological prowess, sound integrative bases of cardiovascular physiology have yet to be elucidated in a large fraction of the human population.
The purpose of the present study was to comprehensively assess the impact of sex and age on the primary physiological determinants of cardiorespiratory fitness and their intrinsic circulatory relationships. To this end, transthoracic echocardiography, central hemodynamics, pulmonary O 2 uptake at rest and during submaximal and peak exercise, as well as intravascular volumes were assessed in healthy young and older women and men matched by age group and physical activity levels.

Study participants
Eighty-five adult women and men matched by age group (young: ≤ 44 years; older: ≥ 50 years) and moderate-tovigorous physical activity (MVPA) levels were recruited via advertisements on community notice boards in the city of Calgary. Inclusion criteria comprised healthy status, absence of current medical symptoms or medication limiting incremental exercise testing, and no history of cardiac, pulmonary or neuromuscular diseases. Individuals fulfilling the above criteria but having donated blood within 3 months prior to the study were excluded. The study was approved by the Conjoint Health Research Ethics Board (REB18-1654) of the University of Calgary and conducted in accordance with the declaration of Helsinki. Prior to the start of the experiments, informed oral and written consents were obtained from all participants.

Experimental design
Participants were required to report once to our laboratory for testing in addition to a voluntary familiarization visit. All individuals were instructed to avoid strenuous exercise, alcohol and caffeine from 24 h prior to testing, as well as to maintain their usual baseline activity and daily dietary habits throughout the study. All measurements were performed after a fasting period (≥ 5 h) in a quiet room with controlled temperature between 22 and 23 ℃. Prior to testing the participants completed demographic and clinical questionnaires and rested in supine position for 20 min in order to stabilize cardiovascular, hemodynamic and hematological variables. MVPA levels were determined from established questionnaires as previously described [26].

Hemoglobin mass (Hb mass ) and blood volumes
Hemoglobin mass (Hb mass ) was determined using the classic carbon monoxide (CO) rebreathing technique integrated in a semi-automated system with a very low typical error of measurement (TE ≤ 1.2%), as previously described [10,27,28]. In brief, following 20 min of supine rest, 2 ml of blood (baseline) was sampled from the median cubital vein via a 20-G venflon (BD, USA) and analyzed immediately in duplicate for percent carboxyhemoglobin (%HbCO), hemoglobin (Hb) concentration and hematocrit (Hct) (ABL80, Radiometer, Denmark). After the exercise test, individuals breathed 100% O 2 for 4 min to flush nitrogen from the airways. After closing the O 2 input, a bolus of 1.5 ml/kg of 99.5% chemically pure CO (Air Liquide, Canada) was administrated into the breathing circuit. Individuals rebreathed this gas mixture for 10 min. Then, an additional 2 ml blood sample was obtained and analyzed in duplicate as aforementioned. The change in %HbCO is used to calculate Hb mass , taking into account the small amount of CO that remains in the rebreathing circuit at the end of the procedure. Total red blood cell volume (RBCV), plasma volume (PV) and blood volume (BV) were determined from Hb mass , baseline Hb concentration and Hct, and normalized by body weight [10,27,28].

Transthoracic echocardiography and central hemodynamics
Apical four-chamber and two-chamber cine-loops were recorded via high-resolution ultrasound (Mindray Medical M9, USA) and analyzed offline (TOMTEC Imaging Systems, Royal Philips, Netherlands) at rest, at a given submaximal workload (100 W), and at peak heart rate (HR peak ) during incremental exercise. Following the American Society of Echocardiography and the European Association of Cardiovascular Imaging recommendations, cardiac chamber quantification was performed using the modified Simpson method (biplane method of disks) by tracing the endocardial border in both apical four-chamber and two-chamber views at end-diastole and end-systole (Additional file 1: Fig. S2) [29,30]. Diastolic function was assessed via transmitral inflow velocities determined by pulsed-wave Doppler, with the sample volume placed between the mitral leaflet tips in the apical four-chamber view. The peak inflow velocity during early (E) and late (A) diastole were assessed, and the E/A velocity ratio was calculated. Moreover, tissue Doppler measurements were performed in the apical four-chamber view, with a 3-mm pulsed Doppler sample volume placed at the level of septal mitral annulus and lateral tricuspid annulus. Early (e') and late (a') peak diastolic annular velocities, peak systolic annular velocity (S'), myocardial isovolumic contraction time (IVCT) and isovolumic relaxation time (IVRT) were determined [29,30]. In parallel, systolic blood pressure (SBP), diastolic blood pressure (DBP) and mean blood pressure (MBP) at the heart level were continuously assessed non-invasively via Finometer PRO (Finapres Medical Systems, Netherlands), with data exported into a pre-established acquisition software (Labchart 7, AD Instruments, UK). SV was determined as LVEDV minus left ventricular end-systolic volume (LVESV), while the product of SV and HR provided cardiac output (Q). Systemic vascular resistance (SVR) was calculated as the ratio of MBP and Q. Echocardiographic variables are commonly presented normalized by body surface area (BSA = 0.007184 weight 0.425 height 0.725 ) to attenuate the impact of body size on sex comparisons [31], except for intrinsic volumetric relationships within the circulatory system (assessed separately in each sex). The reproducibility of key echocardiographic and hemodynamic measurements (withinsubject coefficient of variation (CV)) during incremental exercise in our laboratory is ≤ 6% for left ventricular (LV) volumes and ≤ 3% for blood pressures.

Aerobic capacity
An established incremental exercise protocol [10,27] was implemented using an electromagnetic cycle ergometer (KICKR Core trainer, Wahoo, USA) integrated within a large lower body pressure chamber (LBNP) (165 × 82 × 108 cm) designed for exercise echocardiography. The LBNP comprises an electric hydraulic jack that enables to select any degree from 0 to 45° of left lateral tilting. The combination of the left semilateral body position (17° relative to the horizontal) with lower body negative pressure allows for the simultaneous assessment of cardiac function (which requires a left lateral body position for high-quality and reproducible imaging) and aerobic capacity via the regulation of negative pressure inside the chamber (−50 mm Hg) to induce hemodynamic loads characteristic of the upright position, a physiological requirement to achieve VO 2peak [32,33]. The test started with a warm-up period of 3 min at 20-30 W workloads. Thereafter, the workload was increased by 15-30 W every 60 s until exhaustion was reached in the recommended total duration of 7-10 min. O 2 uptake and CO 2 output were continuously measured and normalized by body weight (CardioCoach VO 2 , KORR Medical, USA). Calibration of the gas analyzers and the flowmeter was performed prior to each test. Breath-by-breath values were averaged over 15 s. The highest breath-bybreath average value was taken as VO 2peak provided that two of the following established criteria were fulfilled: plateau in O 2 uptake despite increased workload, agepredicted HR peak ± 10 bpm [34], respiratory exchange ratio ≥ 1.1 [10,35].

Statistical analysis
Statistical analysis was performed using SPSS 25.0 (SPSS, Chicago, IL). Data were tested for normal distribution with the Kolmogorov-Smirnov test and for homogeneity of variances with the Levene's test. Comparisons among sex and age groups were performed via independent sample t tests. In addition, ANCOVA was used to determine the effects of sex and age on fundamental volumetric relationships in the circulatory system at peak exercise: (i) BV-LVEDV peak and (ii) LVEDV peak -SV peak . A two-tailed P-value less than 0.05 was considered significant. All data are reported as mean (± SD).

General characteristics, resting cardiac structure and function
Main general characteristics of the study participants are shown in Table 1. All individuals were non-smokers and non-obese (body mass index < 30 kg m 2 ). MVPA levels were similar between sex and age groups (P ≥ 0.140). As expected, women presented lower anthropometric indices (height, weight, body surface area) compared with men (P ≤ 0.006), irrespective of the age group. No sex effect was found regarding resting blood pressure and HR (P ≥ 0.302). Nonetheless, older men exhibited higher resting blood pressure (P = 0.014) and older women lower resting HR (P = 0.044) compared with younger sex-matched counterparts. Regarding the cardiac phenotype at rest, sex effects were detected in young individuals, women having smaller LV volumes and SV than men (P ≤ 0.039). Likewise, age effects were noted in the LV, older men presenting lower LVEDV, SV and Q compared with young men (P ≤ 0.038). Furthermore, age per se reduced early peak diastolic inflow and annular velocities (P < 0.001), whereas late peak diastolic inflow and annular velocities were increased with older age (P = 0.001). Accordingly, the E/A velocity ratio was lower in older individuals, irrespective of sex (P < 0.001). Sex effects were specifically detected in older individuals, with women presenting higher tricuspid annular velocities than men (P ≤ 0.045). Furthermore, the myocardial IVCT (septal) was exclusively increased with older age in women (P = 0.001). Table 2 presents key hematological determinants of cardiorespiratory fitness. According to established sex differences, blood O 2 carrying capacity, as represented by Hb concentration and Hb mass , was markedly lower in women compared with men, irrespective of the age group (P < 0.001). An age effect was observed for Hb mass per kg of body weight, which was reduced with older age in men (P ≤ 0.002). With respect to intravascular volumes, all of them (RBCV, PV, BV) were lower in absolute terms in women relative to men, in both young an older age groups (P ≤ 0.016). When normalized by body weight, RBCV remained lower in young and older women compared with age-matched men (P < 0.001), concurring with sex differences in O 2 carrying capacity. With respect to the influence of age, total BV per kg of body weight was reduced in older compared with younger men (P = 0.018).

Sex and age effects at peak exercise
The hematological, hemodynamic, cardiac and pulmonary variables at peak incremental exercise are presented in Table 3. Similar to resting values, blood O 2 carrying capacity at peak exercise was lower in young and older women compared with age-matched men (P < 0.001). Peak blood pressures were similar between sexes (P ≥ 0.075), whereas SBP peak and MBP peak were higher with older versus younger age in women and men (P ≤ 0.018). Higher SVRpeak levels were observed with both female sex (P ≤ 0.038) and older age (P ≤ 0.010). Moreover, HR peak was markedly reduced with older age (P < 0.001). As regards cardiac structure and function, LVEDV peak , SVpeak and Q peak were lower in women compared with men irrespective of the age group (P ≤ 0.046). Furthermore, older age in women and men was associated with reduced Q peak (P ≤ 0.022), primarily explained by lower HR peak since SV peak was not affected by older age (P ≥ 0.595). Absolute peak power output was lower in women compared with men in both age groups (206 ± 34 vs. 308 ± 58 W, P < 0.001 (young); 156 ± 46 vs. 197 ± 46 W, P ≤ 0.046 (older)) and reduced with older age in both sex groups (P < 0.001). Finally, cardiorespiratory fitness, as determined by VO 2peak (mL min −1 kg −1 ), was reduced with older age in both sexes (P < 0.001).  at diastole were noted in older men compared with agematched women (P ≤ 0.005). Moreover, LVEDV 100W was elevated in both older and young men relative to age-matched women (P ≤ 0.049). A sex effect in SV 100W reached significance in younger men, who presented higher SV 100W than age-matched women (P = 0.019).

Sex and age effects at a fixed submaximal workload
No effects of age were observed for submaximal cardiac structure and function (P ≥ 0.108). Concerning submaximal pulmonary variables, absolute O 2 consumption was not affected by sex or age (P ≥ 0.535), whereas ventilation was augmented with older age in women (P = 0.033).

Sex and age effects on cardiovascular volumetric relationships
Figures 1 and 2 present the effects of sex and age, respectively, on fundamental volumetric relationships in the circulatory system, i.e., BV-LVEDV and LVEDV-SV at peak exercise. The slopes of BV-LVEDV peak and LVEDV peak -SV peak did not differ according to sex in each age group (P ≥ 0.062), thus the analysis of the effect of sex on adjusted means was subsequently performed (Fig. 1). A unique sex effect was detected for the adjusted means analysis of BV-LVEDV peak in older individuals, demonstrating a lower LVEDV peak for a given BV in women relative to men (P = 0.024). With respect to the effect of age, the slopes of BV-LVEDV peak and LVEDV peak -SV peak were similar in young and older individuals, irrespective of sex (P ≥ 0.078) (Fig. 2). The subsequent analysis of adjusted means in each sex group did not detect age effects (P ≥ 0.087). Data are expressed as mean ± SD DBP peak , peak diastolic blood pressure; Hb peak , peak hemoglobin concentration; Hct peak , peak hematocrit; HR peak , peak heart rate; LA peak , peak left atria; LVEDV peak , peak left ventricular end-diastolic volume; LVEF peak , peak left ventricular ejection fraction; MBP peak , peak mean blood pressure; O 2 , oxygen; Q peak , peak cardiac output; RA peak , peak right atria; RV EDA peak , peak right ventricular end-diastolic area; RV ESA peak , peak right ventricular end-systolic area; SBP peak , peak systolic blood pressure; SV peak , peak stroke volume: SVR peak , peak systemic vascular resistance; VO 2peak , peak oxygen uptake *P < 0.05 compared with age-matched women a P < 0.05 compared with sex-matched young individuals

Discussion
The main purpose of this study was to assess the influence of sex and age on key physiological determinants of cardiorespiratory fitness. To this end, hematological, cardiac, hemodynamic and O 2 uptake variables at rest and during exercise were analyzed in a cohort of healthy young and older women and men presenting with similar levels of physical activity. The major findings are: (1) BV adjusted by body weight do not differ between sexes; (2) women present smaller LV dimensions and output at peak exercise compared with men, regardless of age and sex differences in body size; (3) for a given absolute BV, older women exhibit lower LVEDV peak than age-matched men. The present findings, along with those of ancillary nature that are conjointly developed hereunder, denote the interactive effects of sex and age on the fundamental basis of the circulatory system at peak working aerobic capacity.
The capacity to deliver O 2 through the circulatory system, i.e., the main determinant of cardiorespiratory fitness [9,10,13], shows great quantitative variability in humans. Both genotypic and phenotypic variation contribute to the wide range of cardiac and aerobic capacities across the lifespan [9]. Beyond the manifest alterations induced by certain cardiovascular and hematological conditions [17,21,22], sex, age and physical activity levels are well-known to independently modulate key limiting factors of the O 2 transport chain [20]. In this respect, human experimental investigations have demonstrated that not all steps of the O 2 cascade from the atmosphere to the systemic circulation have the same importance [10,36]. Lung structural and functional characteristics, which could in principle limit O 2 transport, are overbuilt in relation to the potential delivery of O 2 via the systemic circulation in healthy individuals exercising in normoxic conditions [37]. In men and presumably in women, Table 4 Hemodynamic, cardiac and pulmonary variables at a fixed submaximal workload (100 W) Significant P values (< 0.05) are highlighted in bold Data are expressed as mean ± SD DBP 100W , submaximal (100 W) diastolic blood pressure; HR 100W , submaximal (100 W) heart rate; LA 100W , submaximal (100 W) left atria; LVEDV 100W , submaximal (100 W) left ventricular end-diastolic volume; LVEF 100W , submaximal (100 W) left ventricular ejection fraction; MBP 100W , submaximal (100 W) mean blood pressure; Q 100W , submaximal (100 W) cardiac output; RA 100W , submaximal (100 W) right atria; RV EDA 100W , submaximal (100 W) right ventricular end-diastolic area; RV ESA 100W , submaximal (100 W) right ventricular end-systolic area; SBP 100W , submaximal (100 W) systolic blood pressure; SV 100W , submaximal (100 W) stroke volume: SVR 100W , submaximal (100 W) systemic vascular resistance; Ve 100W , submaximal (100 W) ventilation; VO 2,100W , submaximal (100 W) oxygen uptake *P < 0.05 compared with age-matched women a P < 0.05 compared with sex-matched young individuals VO 2peak is primarily a function of the heart's ability to pump blood and arterial O 2 content [9,10,13]. The latter factor is determined by a relatively fixed and sex-specific hematological variable, i.e., blood O 2 carrying capacity [24]. In contrast, cardiac pumping capacity is primarily subordinated to the filling level of the circulatory system, a rather modifiable variable [6,13]. The observation of similar BV per kg of body weight between sexes, notably in older individuals, yet substantially lower SV peak in women versus men even after taking into account anthropometric differences suggests a potential novel sex dimorphism in a fundamental circulatory relationship [13][14][15]. In both young and older individuals, women exhibit nearly one-fifth lower SVpeak that is paralleled by a commensurate reduction in LVEDV peak compared with men, i.e., the peak filling of the LV is substantially curtailed by the female sex. Integrative cardiovascular factors that may explain such a prominent sex dimorphism in the limitation of cardiorespiratory fitness are discussed hereunder.
The relationship between BV and LVEDV peak reflects the extent to which the distribution of blood is optimized within the circulatory system at peak working capacity. In older individuals, LVEDV peak per unit of BV was lower in women compared with men (Fig. 1). Sex or age effects were not observed in the relationship between LVEDV peak and SV peak , indicating that the Frank-Starling mechanism was preserved. Thus, lower cardiac filling for a given BV may, at least in part, explain the reduced output (Q peak ) in older women relative to age-matched men. Multiple cardiovascular regulatory mechanisms that primarily determine cardiorespiratory fitness might influence venous return and cardiac filling, ultimately limiting the systemic delivery of blood and thereby O 2 [9]. These potential 'weakest' links can be in principle classified as belonging (intrinsic/central) or not (extrinsic/peripheral) Fig. 1 Influence of sex on BV-LVEDV peak and LVEDV peak -SV peak relationships in young and older individuals. Each graph includes the lines of best fit (95% confidence interval). BV, blood volume; LVEDV peak , left ventricular end-diastolic volume; SV peak , peak stroke volume to the anatomy and physiology of the heart. With respect to central factors, it is tempting to hypothesize a role for women's predisposition to cardiac stiffness and impaired LV relaxation with advanced age [16,21,22,38]. Women commonly present with augmented LV diastolic and systolic stiffness and steeper age-related increases in LV systolic stiffness compared with men [38]. Main LV functions may thus be prominently impaired in women with advance age, since older age per se (including men) did not alter the fundamental volumetric relationships (BV and LVEDV peak , LVEDV peak and SVpeak) (Fig. 2). On the other hand, despite older age was associated with an impairment of common markers of LV diastolic function, they did not differ between sexes at rest (Table 1). Indeed, older women had similar mitral but higher tricuspid diastolic annular velocities at rest than age-matched men. While accurate non-invasive measures of diastolic function could not be obtained during exercise due time constraints for multiple cardiac imaging recordings at a given intensity, a prevailing association between resting and peak exercise diastolic function assessments has been found in catheterization studies [39,40]. As regards peripheral factors, skeletal muscle arterial vasodilation is known to independently contribute to increase SV during exercise via the reduction of the systemic resistance to blood flow, i.e., SVR [41]. In this respect, we observed markedly augmented SVRpeak with female sex and older age. In addition, the myocardial isovolumic contraction time (IVCT) was increased with older age in women but not in men, which could also be partly related to Fig. 2 Influence of age on BV-LVEDV peak and LVEDV peak -SV peak relationships in women and men. Each graph includes the lines of best fit (95% confidence interval). BV, blood volume; LVEDV peak , left ventricular end-diastolic volume; SV peak , peak stroke volume female-specific LV systolic and central arterial stiffening [38]. Taken together, older women may have experienced the highest resistance to perfuse the systemic circulation at peak exercise, possibly resulting in lower output (Q peak ), venous return and cardiac filling as a downstream consequence [42], which inverts the aforementioned causality sequence in the circulatory system [43]. Parenthetically, beyond the classic causality dilemma, whether a given factor within an intricate biological mechanism such as the circulatory system is conceptually considered as a cause or consequence depends on the standpoint. Experimental studies manipulating single variables at a time are needed to elucidate why older women exhibit higher SVR peak and distribute their BV in a less optimal manner than opposite-sex counterparts. Such efforts might also shed light into resembling and highly prevalent pathophysiological alterations underlying deathly cardiac conditions in older women [21,38,44,45]. A strong prognostic factor of cardiovascular health, Q peak , was independently modulated by sex and age [7,8]. Specifically, both young and older women presented with reduced Q peak compared with men matched by age, physical activity levels as well as aerobic capacity. In fact, notwithstanding the relevance of cardiac pumping capacity as a hallmark of cardiorespiratory fitness, its pulmonary correlate, i.e., VO 2peak , did not differ between sexes. Likewise, despite cardiac output at a fixed submaximal workload (Q 100W ) was similar in women and men, submaximal O 2 uptake (VO 2100W ) was higher in the former, regardless of age. Collectively considered, for a given unit of systemic blood flow (Q peak , Q 100W ), women extracted more O 2 than men, albeit the female sex is typically characterized by reduced arterial O 2 content owing to lower blood O 2 carrying capacity throughout the adult lifespan [23], as herein demonstrated at rest and during exercise. The increased VO 2 /Q ratio in women may be underlined by central and peripheral factors. Starting from the periphery, recent studies have evidenced higher mitochondrial content and oxidative capacity, theoretically reflecting an enhanced potential for O 2 extraction, in skeletal muscle fibers of women compared with men matched by VO 2peak [46][47][48]. Furthermore, the current observation of higher SVR peak in women could also reflect an increased tightening of the systemic circulation leading to improved blood flow distribution and enhanced O 2 extraction at peak exercise [49]. Yet, at a fixed submaximal workload another factor could be at play. Women have smaller conducting airways than men, also when adjusted by lung size [50]. Consequently, the female lung's anatomy imposes greater mechanical resistance to breathing, particularly during exercise, thereby entailing a higher ventilatory work and O 2 uptake (energy expenditure) for a given external work output [50]. The present findings therefore support the notion of women being less physiologically efficient than men, even when crucial counfounders are matched between sexes.
There are some limitations in this study that require comment. As an intrinsic limitation of cross-sectional analyses assessing the influence of sex and age, causeand-effect relationships cannot be established, thus the conclusions should be taken with caution. Very longterm (lifelong) longitudinal studies, while seldom implemented due to outstanding logistical and technical challenges [51], could eventually provide a stronger level of evidence. Moreover, we selected healthy individuals in order to limit the influence of disease-related confounding factors. Further research will elucidate whether the present findings can be extrapolated to specific clinical conditions.

Perspectives and significance
The current study assessed the impact of sex and age on major determinants of cardiorespiratory fitness in a cohort of healthy individuals matched by sex, age and physical activity levels. Following a comprehensive analysis at rest and during exercise, substantially smaller LV dimensions and correspondingly lower cardiac pumping capacity in women relative to men, independently of age and body size, are consistently demonstrated. Moreover, sex and age interact on a fundamental circulatory relationship, i.e., that of total circulating BV with LVEDV. Older women require higher BV for a certain level of LV filling, which suggests the presence of an intrinsic cardiovascular constraint, which concurs with functional features of prevalent life-threatening conditions in this population. Further experimental studies will ascertain whether such a circulatory limitation specific to older women can be overcome by tailored lifestyle and/or pharmacological interventions.