Individuals were recruited at the National Institute of Integrative Medicine (NIIM; Melbourne, Australia) between October 2020-February 2024. Recreational endurance athletes (aged 40-65 years) were invited to participate in the present double-blind parallel randomized placebo-controlled dose-response trial of 12 weeks duration investigating the effect of Kyolic aged garlic extract on arterial stiffness and aerobic fitness during high intensity exercise on a cycle ergometer. Recreational endurance athletes were defined as regularly undertaking ≥3 30 min/week moderate to high intensity exercise sessions, such as triathlon, cycling, running or swimming.

The American Heart Association defines moderate exercise intensity at 50-70% of the maximum heart rate and vigorous/high exercise intensity at 70-85% of the maximum heart rate (39). Individuals were recruited through local fitness and masters cycling clubs, the NIIM newsletter and NIIM social media.

Individuals were excluded from the study if they were diagnosed with a cardiovascular condition, including a previous heart attack or stroke, or had a chronic illness, such as cancer. Additionally, individuals were also excluded if they had a recent surgery or were planning any surgery or change in medications, were pregnant during study or reported an intolerance to garlic.

Participants were instructed to cease any garlic supplementation prior to enrolment, and not to change their exercise routine or diet during the 3 month trial. As Kyolic aged garlic extract is consumed as a supplement with distinctly different therapeutic effects compared with raw or cooked garlic, the dietary intake of garlic was allowed to remain unchanged.

The study was conducted in two parts, the first cohort was recruited between October 2020 and January 2022 and the second cohort was recruited between January 2023 and February 2024. Consenting eligible participants in the first cohort were randomly allocated to the low dose consisting of two capsules of aged garlic extract [Kyolic^® Reserve Formula (Wakunaga of America, Co., Ltd.,) containing 1.2 g AGE powder and 1.2 mg S-allylcysteine (SAC)] or placebo (containing inert microcrystalline cellulose) daily for 12 weeks. In the second cohort participants were randomly allocated to the high dose consisting of four capsules of aged garlic extract (Kyolic^® Reserve Formula containing 2.4 g AGE powder and 2.4 mg SAC) or placebo taken daily for 12 weeks. A computer-generated permuted random number table provided by an independent researcher not involved in recruitment, data collection or follow-up for the present study was used for randomization.

Kyolic aged garlic extract powder was manufactured from organically grown garlic bulbs, which have undergone a 20 month natural ageing process at room temperature. Placebo capsules were matched in appearance to the opaque active capsules and packaged in identical containers by the manufacturers off-site. Activated carbo-sachets were added to each container to disguise any odor. Participants, investigators and researchers were blinded to treatment allocation. Blinding success was assessed at the end of the trial by questionnaire.

Participants were instructed to start taking their trial capsules with food to minimize belching after the first VO₂max testing session at baseline. Compliance was assessed by questionnaire and by capsule count throughout the study.

Participants had four monthly appointments at NIIM (weeks 0, 4, 8 and 12), and two appointments at the Mets Performance exercise physiologist center (Melbourne, Australia) for testing of aerobic capacity with VO₂max on a cycle ergometer, both at the beginning (week 0) and at the end of the study (week 12).

Participants' age, sex, BMI, regular medication and supplement intake and exercise routines were collected at baseline.

The participants' exercise routines were assessed using an online questionnaire adapted from the International Physical Assessment Questionnaire (40) to quantify the length of regular moderate and high intensity exercises throughout the week.

Arterial stiffness was assessed by non-invasive validated methods measuring PWV (41) using the SphygmoCor Xcel (Cardiex) (11) in cohort 1 and the Mobil-O-Graph device (IEM GmbH) (12) in cohort 2, due to the SphygmoCor Xcel no longer being supported in Australia from 2023. The Mobil-O-Graph has been validated against the gold standard tonometry for central blood pressure (14) and pulse wave analysis (42). However, PWV measurements are not directly comparable, whereby the Mobil-O-graph measured consistently lower PWV values compared with the SphygmoCor tonometer (43), therefore the PWV was presented separately for each cohort. The SphygmoCor Xcel (11) measures PWV directly by tonometry from carotid to femoral artery, while the patient is in a supine position. The Mobil-O-Graph (12) is a non-invasive device measuring central hemodynamic measures via a blood pressure cuff, such as blood pressure, pulse pressure, heart rate and PWV, an indicator for arterial stiffness. PWV is the speed in which the blood travels through the arteries in m/sec and is derived from the brachial pulse and indirectly calculated by a computerized algorithm and compared with age- and sex-adjusted population references values.

Participants were advised to refrain from drinking coffee or having a large meal for at least 2 h prior to the measurements and after at least 5 min of rest. Blood pressure was taken using a calibrated and validated digital sphygmomanometer (Omron HEM-907; Omron Healthcare Singapore Pte Ltd) (44).

Aerobic fitness was measured by VO₂max and aerobic and anaerobic/lactate threshold during high-intensity exercise using a cycle ergometer test station. The cyclist was connected to an oxygen analyzer, which measured the ratio of CO₂ to the volume of oxygen intake. Blood lactate concentration was measured using a finger prick throughout the exercise workout.

VO₂ max is defined as the maximum amount of oxygen your body can take in, transport via the blood stream into the muscles and utilize in 1 min.

Relative VO₂ max (ml/kg/min) measures the maximum amount of oxygen that an individual can utilize during intense or maximal exercise, in ml/kg body weight/min. Absolute VO₂ max (ml/min) is the total volume of oxygen regardless of body weight and is important for non-weight bearing activities such as swimming and flat cycling.

The exercise workloads were selected to gradually progress in increments from moderate to maximal intensity. In the present study workload in increments of 30 Watts of power was increased every 3 min, starting at 60 Watts (Fig. 1A).

When exercising, energy is first produced aerobically by using oxygen and primarily by burning fat and carbohydrates, such as glucose. The aerobic threshold is reached when the production of energy starts to become dominated by anaerobic glycolysis (sugars) rather than the oxidation (aerobic in nature) of fats. Anaerobic energy production without oxygen produces lactate or lactic acid, and the lactate threshold is defined as the level of exercise intensity causing lactate to accumulate in the blood at a faster rate than it can be removed (Fig. 1B).

The lactate threshold test consisted of multiple blood lactate concentration measurements by a finger prick at the immediate completion of each incremental workload (30 Watts) during the cycling test, which was conducted until the participant's maximal intensity was reached. The lactate threshold occurs at a percentage of an athlete's VO₂max based on their training status. In untrained individuals the lactate threshold is reached at ~50-60% of VO₂max and in trained individuals at ~70-80% of VO₂max. A higher lactate threshold correlates to better overall athletic performance.

Recovery time was assessed in days by a simple questionnaire, and muscle soreness by a pain 10-point Likert scale at baseline and at 12 weeks.

Urinary proteomic analysis was carried out by the Commonwealth Scientific and Industrial Research Organisation Land and Water following the protocol published by Zimmerli et al (35), featuring a panel of 22 cardiovascular risk-relevant peptides using high-resolution liquid chromatography-mass spectrometry (LC-MS). Proteins (>3 KDa) were removed from urine samples using Amicon Centrifugal Filter Units at 16,100 x g at room temperature for 5-10 min using an Eppendorf centrifuge. Urine samples were normalized to the measured creatinine concentration and urine samples were dried in a speed-Vac. Sample pellets were resuspended in 40 ul of 0.1% formic acid in preparation for LC-MS analysis and 5 ul was injected onto the LC-MS for analysis. Samples were desalted and concentrated with a trap column (PepMap100 C18 5 mm x 300 µm, 5 µm, Thermo Scientific. Inc.) and separated on a nano column (PepMap100 C18 150 mm x 75 µm, 2 µm, Thermo Scientific. Inc.) using an Ultimate^TM 3000 RSLC nano LC system (Thermo Scientific. Inc.). Flow rate was 300 nl/min and system temperature was 35˚C. Mobile phase A consisted of water and 0.1% (v/v) formic acid and mobile phase B consisted of 80% (v/v) acetonitrile and 0.08% (v/v) formic acid. Urine peptides were eluted using a gradient of 5 to 60% solvent B for 20 min and 60 to 99% solvent B for 7 min. The eluted peptides were ionized with a Nanospray Flex Ion Source (Thermo Scientific. Inc.). The spray voltage was set to 2.3 kV and the temperature of the heated capillary was set at 300˚C. Sweep nitrogen gas flow rate was at 1.5l/min and source pressure was at 1.48 Torr. After ionization, mass spectra (MS1) and tandem mass spectra (MS/MS) analysis was performed using an Orbitrap Fusion MS (Thermo Scientific. Inc.). MS survey scans of peptide precursors were performed in the Orbitrap detector and the scan range was 400 to 1500 m/z at resolution of 120 K (at 200 m/z). The target value of automatic gain control (AGC) was set as 4 x 10⁵. The maximum injection time for the MS was 50 ms. MS/MS was performed on the most abundant precursors of charge states 2+ to 7+ with intensity greater than 1 x 10⁵. They were isolated by the quadrupole with a window of 1.6 m/z. Fragmentation was achieved by high-energy collisional dissociation (HCD) with collision energy of 28%. Fragments were detected in the ion trap detector in rapid scan rate mode. The AGC target was 4 x 10³, maximum injection time was 300 ms and the dynamic exclusion was 15 sec. The instrument was set to run in top speed mode with a three sec cycle for both the MS and MS/MS scans.

A list of 81 urinary peptides identified by Brown et al (36) was used to search peptides in the urine samples. Precursor mass tolerance was set to 10 ppm and product ions were searched at 0.6 Da. No-Enzyme (unspecific) was selected. Modification included Oxidation (+15.995 Da). Creatinine, a clearance protein, was measured to adjust for differences in urinary concentration between patients due to hydration variances. Urine creatinine concentration was measured using an ELISA-based assay using a Creatinine Urinary Detection Kit (cat. no. EIASCR; Invitrogen; Thermo Fisher Scientific, Inc.) and measured using a UV-Vis spectrometer at a wavelength of 490 nm. Each urine sample was diluted 1:20 and 50 µl was used for the assay as per the manufacturer's protocol.

Primary outcome measures were aerobic fitness, including VO₂max, aerobic and lactate threshold and power. Secondary outcome measures were recovery time, muscle soreness, PWV and proteomic biomarkers. Analyses were performed using SPSS Inc. (version 26; IBM Corp.). P<0.05 was considered to indicate a statistically significant difference.

Differences between groups at baseline were assessed by χ² test or Fisher's Exact test if the values were <5 for categorical variables and by an unpaired Student's t-test for continuous variables.

The participant's mean change score (value at 12 weeks- value at baseline) was calculated for each outcome measure and ANCOVA with Bonferroni's correction was performed. Relevant baseline values were used as co-variates to assess the mean differences between groups (garlic vs. placebo).

For the main statistical analyses, cohort 1 and cohort 2 were combined (low and high dose), and subgroup analysis was performed by cohort.

Polypeptide data were normalized to creatinine clearance protein. Missing values were replaced by one-fifth of the minimum positive values of their corresponding variables. The acquired data were then log10 transformed and scaled using the mean-centred value and divided by the SD of each variable. Data were then subjected to univariate and multivariate statistical analysis using SIMCA (version 17.01; Sartorius Stedim Biotech; Sartorius AG). GraphPad Prism (version 9.3.0; Dotmatics) was used for descriptive statistical analysis and plot generation.

A 15 peptide quality assurance and quality control (QAQC) standard consisting of 15 heavy isotope-labelled peptides (cat number 88321; Thermo Fisher Scientific, Inc.) was analysed throughout the analytical sequence and was found to be with a 5% relative standard deviation, demonstrating good analytical performance of the column and MS detector.

The trial was conducted in two parts whereby the first cohort, allocated the low trial supplement dose, was recruited between October 2020 and January 2022 during the COVID-19 lockdowns and the second cohort, allocated the high trial supplement dose, was recruited between January 2023 and February 2024.

A total of 108 potential participants were screened for eligibility, with a total of 24 either not fitting inclusion criteria (n=18) or losing interest shortly after enrollment into the study (n=6). Of the potential participants, one was not eligible on medical grounds due to a high PWV reading at screening prompting a consultation with a cardiologist and a subsequent diagnosis of severe coronary artery disease, despite not having any other risk factors such as hypertension or high cholesterol. A further nine participants withdrew after baseline VO₂max testing, either due to losing interest in the study (n=4), the long travel distance to and from the study site (n=2), COVID-19 lockdown (n=1), study-unrelated gastroenteritis (n=1) or medical advice from the patient's doctor (n=1). In total, 75 participants completed the study, including 37/40 (92%, cohort-1) and 38/40 (95%, cohort-2). The mean age was 53±8 years and two-thirds of the participants were male (Table I).

All participants in the first cohort were serious endurance athletes who trained an average of 1.5 h daily with moderate to high intensity exercise (~1 h moderate and ~30 min high intensity) of either cycling, running or swimming. The majority (86%) of participants in cohort-1 were cyclists, cycling up to 400 km or 17 h per week (Table II).

The second cohort was significantly less athletic overall compared with the first cohort, training an average of 45 min daily of moderate to high intensity exercise (~30 min moderate and ~15 min high intensity) (Table II). Participants in cohort-2 exercised through a variety of sports including cycling, running, rowing, swimming, kick boxing, Bikram hot yoga and underwater hockey (Table II, Fig. 2).

At baseline there were no statistical differences in blood pressure or PWV between the garlic vs. placebo groups. However, at the end of the study, a marked improvement in blood pressure and PWV was observed in the Kyolic group compared with the placebo group in a subgroup of participants with elevated blood pressure (>135 mmHg SBP), as well as in PWV (Table III).

Kyolic aged garlic extract significantly improved aerobic fitness, as evidenced by increased absolute VO₂max (mean difference, 133±83 ml/min; P<0.07), power (mean difference, 10.2±4.9 W; P<0.04), lactate threshold to power output (mean difference, 11.8±3.2 W; P<0.001), lactate threshold to oxygen uptake (mean difference, 174±72 ml/min; P<0.02) and aerobic threshold to power output (mean difference, 9.1±4.5 W; P<0.045) (Table IV, Figs. 3 and 4).

While there were no significant differences between the garlic and the placebo groups in participants' types of sports (Table II, Fig. 2), there were differences in exercise regimes and fitness levels between the first cohort (low dose) and the second cohort (high dose) at baseline. While the first cohort exercised a mean of ~90 min/day (54 min of moderate intensity and 40 min of high intensity/day), the second cohort exercised a mean of 45 min/day (30 min moderate and 15 min high intensity/day), which was approximately one-half of the time exercising and at a lower intensity compared with the first cohort (Table V).

As the baseline levels of certain aerobic fitness measurements, such as VO₂max, were significantly different between the two cohorts (low dose vs. high dose, P<0.001), a dose-response analysis was not feasible (Table V). The baseline fitness of the two cohorts would have influenced or confounded any dose-dependent outcome measures. However, in both cohorts, the Kyolic group had statistically significant improvements in aerobic fitness compared with the placebo group (Table SI).

Compared with cohort 2, aerobic capacity improved significantly in cohort 1, as evidenced by increased absolute VO₂max (mean difference, 238±105 ml/min; P<0.03), power (mean difference, 16.2±7.3 W; P<0.035), lactate threshold to power output (mean difference, 16.9±5.4 W; P<0.003), lactate threshold to oxygen uptake (mean difference, 269±104 ml/min; P<0.014) and aerobic threshold to oxygen uptake (mean difference, 245±90 ml/min; P<0.01; Table SI).

Kyolic aged garlic extract significantly improved aerobic fitness, as evidenced by quicker recovery times compared with the placebo group in both treatment cohorts combined (n=70; mean difference, -0.6±0.3 days; P<0.05) (Table VI, Fig. 5). Additionally, the difference in the recovery time was more pronounced in the cohort 1 (low dose), with the garlic-treated group recovering significantly quicker compared with the placebo group after exercise (mean difference, -1.1±0.5 days; P<0.025) (Table VI). Muscle soreness was reduced in both groups compared with baseline (albeit not significantly; Fig. 6).

In a subgroup of participants from the first cohort (n=22) and the second cohort (n=37), a panel of 22 polypeptides and their derivates were tested. As higher concentrations of these polypeptides have been associated with higher cardiovascular risk, a reduction in the levels of these polypeptides is desirable. It was demonstrated that a subset of polypeptides significantly decreased in the Kyolic group compared with the placebo group, leading to an improved cardiovascular risk proteomic profile in the Kyolic group at the end of the study (Table VII, Fig. 7). Proteomic analysis differed between the cohorts. In the low dose cohort, 10/22 (45%) of the peptide concentrations were significantly reduced in the Kyolic group compared with the placebo (Fig. 7a) and 3/22 (13%) were significantly reduced in the second cohort (Fig. 7b). Furthermore, the mean polypeptide baseline levels varied between the cohorts. Therefore, a combined proteomic analysis of both cohorts was not statistically meaningful.

Compliance was satisfactory, with all participants taking their trial supplements >95% according to protocol.

Tolerability was high, with 1 participant in the placebo group noting a weight gain of 2-3 kg over the course of the study and 1 participant in the garlic group noting a lower heart rate. In the garlic group, 1 participant reported burping, which did not bother her and another participant noted that her menstrual cycle improved becoming more regular and less painful.

Blinding was successful in both cohorts. In the low dose cohort 1, the majority (80%) were unsure of their group allocation, while 5% guessed their group allocation incorrectly, verifying successful blinding. In the high dose cohort 2, a greater proportion guessed correctly (37%) compared with the low dose cohort 1 (16%) and two-thirds (63%) were unsure or incorrect (Table VIII).