Cardiovascular diseases (CVD) are the leading causes of non-communicable deaths worldwide (WHO, 2018). Although the consequences of CVD, such as a myocardial infarction or stroke, do not manifest until adulthood, the pathobiology of CVD begins in childhood (Juonala et al., 2005). The atherosclerotic process is progressive and is associated with the development of CVD risk factors during childhood as well as early deterioration in vascular function (Aggoun et al., 2005). Promoting cardiovascular health in childhood and identifying those with elevated risk is therefore important for the prevention of CVD outcomes in adulthood (Urbina et al., 2009).
Aerobic fitness is a strong predictor of cardiovascular health with low aerobic fitness being associated with elevated CVD risk in children and adolescents (Barker et al., 2018; Ekelund et al., 2007; Steele et al., 2008). Conversely, a high level of aerobic fitness has been associated with lower blood pressure (Nielsen and Andersen, 2003), improved lipid profiles (Hager et al., 1995), lower levels of inflammatory markers (Isasi et al., 2003), lower rates of obesity (Janz et al., 2002) and enhanced insulin sensitivity (Lee et al., 2006) in children and adolescents. More recently, aerobic fitness during adolescence has also been found to relate to myocardial infarction (Hogstrom et al., 2014) and mortality (Hogstrom et al., 2016) in adult life. These data suggest a possible protective effect of attaining high levels of aerobic fitness within youth upon CVD risk factors and vascular and haemodynamic function.
Exercise training has been shown to enhance both macro and microvascular function in adults (Green et al., 2011) and adolescents (Bond et al., 2015). Data in healthy children are sparse, but an association between aerobic fitness and macrovascular function has been shown in 6 to 11 year old children (Agbaje et al., 2019; Hopkins et al., 2009; Reed et al., 2005; Veijalainen et al., 2016). Sakuragi and colleagues (2009) also reported a negative association between aerobic fitness and arterial stiffness in 573 10-year olds, although the association became weaker after adjusting for adiposity. Aerobic fitness was assessed indirectly using the 20 m shuttle run test, which although correlates with a direct measurement of peak oxygen uptake (V[O.sub.2] peak) in children (Castro-Pinero et al., 2010), the explained variance can be less than 50% (Mayorga-Vega et al., 2015). Furthermore, the test is strongly influenced by adiposity (Olds and Dollman, 2004) and has been criticised for being a measure of performance and thereby its use as a prediction for estimating peak V[O.sub.2] has been questioned (Armstrong, 2018). Thus, the recommended 'gold-standard' measurement of aerobic fitness requires gas analysis during maximal exercise to directly measure V[O.sub.2] peak (Barker et al., 2013) with adjustment for body size and composition (Loftin et al., 2016). Nadeau and colleagues (2010) directly measured V[O.sub.2] peak and reported a positive correlation with forearm blood flow in healthy adolescents and adolescents with type I diabetes. Although the group directly measured V[O.sub.2] peak, it was expressed as a ratio-standard for body mass (mL.[kg.sup.-1].[min.sup.-1]) which may not adequately control for body size (Welsman and Armstrong, 2000). Allometric scaling methods are recommended to control for the influence of body size on aerobic capacity and have been shown to reduce the size of the relationship between aerobic fitness and markers of CVD risk in children when compared to the ratio standard method and after accounting for body composition (Agbaje et al., 2019).
Microvascular dysfunction is an important predictor of early atherosclerosis (Wilkins et al., 2012) and provides additional valuable information on vascular risk. However, our understanding of the relationship with fitness and microvascular function in paediatric groups is limited. There is evidence to suggest more physically active children have better retinal arteriolar caliber (Gopinath et al., 2011) but not microvascular function (Radtke et al., 2013). Reed and colleagues (2005) identified that healthy children with higher levels of aerobic fitness had improved small vessel function, although only arterial stiffness was assessed and V[O.sub.2] peak was estimated from the 20 m shuttle run test. Paediatric analyses of microvascular function have typically included overweight or obese children (representing > 50% of the sample) (Bastos da Cunha et al., 2017; Hedvall Kallerman et al., 2014) or those with type I diabetes (Roche et al., 2008). Furthermore, studies that have focused on the association between microcirculatory function and aerobic fitness typically use a single measure of vascular function (Reed et al., 2005) or quantify aerobic fitness using the ratio standard expression for V[O.sub.2] peak without appropriate statistical justification (Roche et al., 2008).
The aim of this study was to examine differences in macro and microvascular function between 9-10 year old children stratified for 'higher' and 'lower' directly assessed aerobic fitness (i.e. V[O.sub.2] peak) with appropriate adjustment for body size and composition. We also explored differences in CVD risk factors and blood markers including insulin resistance between levels of aerobic fitness. We hypothesised that: 1) children with higher aerobic fitness will have superior macro and microvascular function in comparison to those with lower aerobic fitness, even after adjusting for adiposity and 2) children with lower aerobic fitness will have a poorer CVD risk factor profile including increased insulin resistance compared to those with higher aerobic fitness.
Children from three state primary schools in the South West of England were invited to participate in the study and 100 children (53 males, 47 females) aged nine and 10 years volunteered. One girl withdrew following her initial visit and three girls were removed from the study due to use of vasoactive medication leaving a total sample of 96 children (53 males, 43 females). Written informed assent and consent was obtained from the child and their parents/guardians respectively. A medical screening questionnaire was completed and signed by the parents/guardians. The Local Medical Research Ethics Committee provided ethical approval for the study.
The children visited the University laboratories on three occasions over a 1-month period where practical. During the first visit anthropometric measures and body composition were assessed. During the second visit a maximal exercise test was completed and during the third visit vascular function and blood markers were assessed.
Measurements were performed while the child was wearing either a lightweight t-shirt and shorts or a swimming costume with no footwear. Stature was measured using a stadiometer to the nearest 0.01 m. Body mass was assessed with a balance beam scale and recorded to the nearest 0.1 kg. Body mass index (BMI) was calculated. Children were classified as lean, overweight or obese using age and sex-specific BMI cut-offs (Cole et al., 2000).
Total body fat percentage (TBF%) was calculated following air displacement plethysmography (BOD POD 2000A). Body volume was adjusted for lung volume using tidal volume estimates. TBF% was calculated from body density using Lohman's child adjusted Siri equation (Lohman, 1989). TBF% was obtained in 86 of the participants and subsequently used to estimate fat free mass (FFM).
Visceral adipose tissue (VAT) was assessed using magnetic resonance imaging. Examinations were performed using a 1.5 Tesla superconducting magnetic resonance scanner (Gyroscan Intera, Philips). Interleaved transverse slices (8 mm thick, with a 1 mm gap between them) were acquired from lumbar region (L)1 to L5 with a typical voxel size of 2.5 x 3.5 x 8.0 mm. Scanning was performed with the participant performing a ~ 20 s breath hold to minimize movement artifacts. Data are reported for 86 children for whom successful scans were obtained. The estimated volumes for VAT and the total cavity from L1 to L5 in c[m.sup.3] were determined.
Peak V[O.sub.2] was determined during a ramp exercise test to volitional exhaustion on an electro-magnetically braked cycle ergometer (Excaliber Sport; Lode). Heart rate (HR) was measured continuously using HR telemetry (Polar Vantage NV, Polar Electro Oy). Gas exchange variables (e.g. V[O.sub.2]), were measured continuously using an online breath-by-breath metabolic cart (Cortex Metalyzer 3B; Cortex Medical) which was calibrated prior to each test. The test began with a 5-minute warm up at 20 watts (W), followed by a 10 W increment per minute protocol until voluntary exhaustion, despite strong verbal encouragement. Participants were asked to maintain a pedal cadence of ~ 70 rpm throughout the test. Peak V[O.sub.2] was taken as the highest 10 s average V[O.sub.2] during the test, which in our labor atory results in the measure ment of a true maximal V[O.sub.2] in ~ 90% of cases (Barker et al., 2011; Sansum et al., 2019). Absolute peak V[O.sub.2] (L x [min.sup.-1]) was initially ratio scaled to body mass (mL x [kg.sup.-1] x [min.sup.-1]) and FFM (mL x kg [FFM.sup.-1] x [min.sup.-1]). Preliminary analyses revealed a significant residual size correlation for ratio scaling of body mass (r = -0.40; p
Vascular function was assessed in the morning following an overnight fast and conducted in a quiet, temperature-controlled room (22.0 [+ or -] 0.5 [degrees]C) with the child lying in the supine position following an acclimatisation period of 30 minutes.
Pulse wave analysis
Applanation tonometry of the radial pulse of the right arm enabled analysis of the peripheral waveforms. A high-fidelity micromanometer (SphygmoCor[R]; AtCor Medical Pty. Ltd., Sydney, Australia) was used to flatten, but not occlude, the...