Efficacy of a nootropic spearmint extract on reactive agility: a randomized, double-blind, placebo-controlled, parallel trial

A randomized, double-blind, placebo-controlled, parallel design was implemented to determine the effects of PSE on reactive agility in healthy individuals aged 18–50 years. A total of 485 individuals were initially contacted via phone to screen for eligibility (Fig. 1). Individuals who met inclusion/exclusion criteria were asked to report to the laboratory (MusclePharm Sports Science Institute, Denver, CO) for a familiarization visit so that the reactive agility tests could be practiced to minimize any learning effects (Fig. 2). One hundred forty-two participants were randomized into PSE or placebo (PLA) groups in a staggered start design (1 week after screening visit). To generate a randomization schedule using block randomization, participants were stratified by sex and age into four groups: young (18–35 years)/male, older (36–50 years)/male, young/female, and older/female. A random sequence of blinded treatments (A and B) was generated by the principal investigator using 37 blocks of 4 on a research randomizer website (www.​randomizer.​org). These blocks were applied to each stratum mentioned above as each previous block of 4 was filled by participants. All blood was collected at a LabCorp facility and all other measures were performed at the MusclePharm Sports Science Institute.

At baseline (Day 0) participants were instructed to take two capsules of either PLA or PSE (providing 900 mg PSE/day containing a minimum of 14.5% rosmarinic acid and 24% total polyphenols (expressed as gallic acid equivalents); Neumentix™ Phenolic Complex K110–42, Kemin Foods, L.C., Des Moines, IA) each morning with breakfast for 90 days, except on test days, when they took the capsules directly after testing. Both types of capsules were produced by the same manufacturer to be identical in shape, size, and color, and were sealed in identical bottles. All investigators involved in product dispensing, data collection, and analysis of outcomes were blinded in the following manner: PSE and PLA bottles were labeled “A” or “B” by unblinded individuals at Kemin Foods LC. who were not involved in subject interaction or data assessment. All bottles were labeled according to ICH-GCP guidelines. At each post-supplementation study visit (Days 7, 30, and 90) capsules were counted for compliance. Daily compliance was also logged by the participants in a study diary as a secondary measure. After fasting for 10 h, participants arrived at the laboratory at baseline, and after 7, 30, and 90 days of consuming the supplement daily. Testing consisted of an interactive audio-visual reaction time test battery, a blood draw, and vital signs assessment. After testing, a meal replacement bar was provided to participants (Combat Crunch™; MusclePharm Inc., Denver, CO). An Institutional Review Board (Quorum Review IRB, Seattle, WA) approved the study protocol and informed consent documents prior to initiation of the study. Furthermore, all study procedures were consistent with Good Clinical Practices under the United States Title 21 of the Code of Federal Regulations and the Declaration of Helsinki.

Healthy, recreationally-active men and women between the ages of 18–50 years were recruited for the present study. Healthy was defined as having a body mass index between 18.5–29.99 kg/m² or between 30.0–34.99 kg/m² with a body fat of < 39% for women aged 18–39 years, < 40% for women aged 40–50 years, < 25% for men aged 18–39 years, or < 28% for men aged 40–50 years [42]. Body fat was determined using bioelectric impedance analysis (InBody Co., Seoul, South Korea). Health status was assessed using a health history questionnaire. A recreationally-active individual was defined as ≥1 and ≤ 6 h of moderate and/or vigorous exercise per week as measured by the New Zealand Physical Activity Questionnaire [43]. Eligibility also included willingness to abstain from caffeine-containing products for 10 h prior to and during all test visits, from alcohol consumption and physical activity for the previous 24 h, and from strenuous resistance exercise (defined as heavier than a 6-repetition maximum [the most weight which a participant could successfully lift 6 times]) for 48 h. Participants were required to keep their sleep consistent (within 2 h) each night prior to all study visits. Finally, participants were required to maintain their normal diet (including repetition of dietary intake for the 24 h prior to each study visit), exercise, and sleep regimens throughout the study. All females were amenorrhoeic during test visits. A high school diploma or the equivalent was also necessary for inclusion.

Participants were excluded from the study if they had a history or presence of a clinically important cardiac, renal, hepatic, endocrine (including diabetes mellitus), pulmonary, biliary, gastrointestinal, pancreatic, uncontrolled hypertension, depression or neurologic disorder. Presence of cancer in the past 2 years was a criterion for exclusion. Additional criteria for exclusion included: colorblindness, pregnancy, history of alcohol abuse (12 months), history of repeated minor head injury or a period of unconsciousness of 1 h or more, high consumption of caffeine (≥500 mg/d), marijuana use (in prior 2 months), tobacco use (in prior 6 months), an active infection, use of psychotropic medications (in prior 4 weeks), use of supplements known to improve cognitive function, an allergy to ingredients of the test product or the snack provided, or a sleep disorder or occupation where sleep during the overnight hours was irregular.

The interactive audio-visual reaction time test battery was performed on the Makoto Arena II device (Makoto USA Inc., Elk Grove Village, IL), a device which has been employed in studies previously [40, 44‐47]. The Makoto device is a 360⁰ audio-visual device consisting of 3 towers 1.83 m tall in the shape of a triangle with 2.44 m between towers. Each tower consists of 16 targets: 12 on each tower (levels a, b, and c; see Fig. 3) and 4 on each footplate on the floor (level d; see Fig. 3) intended to be struck with hands and feet, respectively. A line of black tape was placed on the ground whose center is 1.55 m from each of the first two towers. Participants started behind the line and faced forward for all tests. Ten tests were utilized to address a variety of perception and movement modalities. The first three tests involved a single-step protocol, where a single target on one of two towers (tower 1 or 2, level b, top of the diamond of four panels; see Fig. 3) would light up once and the participant would move toward the target and strike it. The three single-step tests utilized either an audio stimulus, a visual stimulus, or a combined audio and visual stimulus. To eliminate guessing order by the participant, the single-step tests were each performed five times in a random order with two tests to the left (tower 3), two tests to the right (tower 1), and one random direction (left or right) with its score discarded. Duplicate scores for each side (left or right) were averaged for reporting purposes, such that each of the three single-step tests resulted in a score for left and a score for right. The remaining seven tests were all timed tests with the panels exhibiting both audio and visual stimuli in unison until the allotted time was completed. The fourth and fifth tests were 30 s stationary tests without or with footplates (level d), in which the participant was in front of one tower (tower 1) and the 12 tower targets (or 16 targets with footplates) were struck as quickly as possible. The sixth and seventh tests were lateral movement tests utilizing two towers (towers 1 and 2), which involved the participant moving back and forth between the two towers as the 12 hand targets from each of the towers (or 16 targets with footplates) activated one at a time for 30 s. The eighth and ninth tests were 30-s multi-directional movement tests using all three towers, and involved the participant moving among the three towers as the 12 hand targets from each of the towers (or 16 targets with footplates). The tenth test was identical to the ninth, except that it lasted 2 min. All the 30-s tests (tests 4–9) were performed in duplicate and scores were averaged. Reaction performance was captured as the number of hits and average reaction time (ART) for all tests, except the single step testing for which only one hit occurred each test and therefore only reaction time was captured. Targets were activated for a maximum of 10 s, so activated targets could not be missed since participants had ample time to strike them. Since the single step tests were short in duration and not fatiguing, the rest periods were as short as possible: 5 s. Between every iteration of each 30-s test, participants rested for 30 s. Prior to the 2-min test, participants were given 60 s of rest.

After a 10 h overnight fast, blood samples were collected via venipuncture by a trained phlebotomist. After collection, samples were analyzed (Laboratory Corporation of America, Burlington, NC) for the following parameters: complete blood count (CBC) [white blood cell count (WBC), red blood cell count (RBC), hemoglobin, hematocrit, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red blood cell distribution width (RDW), platelets (absolute), neutrophils (percent and absolute), lymphocytes (percent and absolute), monocytes (percent and absolute), eosinophils (percent and absolute), basophils (percent and absolute)], comprehensive metabolic panel (CMP) [serum glucose, blood urea nitrogen (BUN), creatinine, estimated glomerular filtration rate (eGFR), BUN:creatinine, sodium, potassium, chloride, carbon dioxide, calcium, protein, albumin, globulin, albumin:globulin (A/G), bilirubin, alkaline phosphatase, aspartate aminotransferase (AST), alanine aminotransferase (ALT)], lipids [total cholesterol, triglycerides, high density lipoprotein (HDL) cholesterol, very low density lipoprotein (VLDL), low density lipoprotein (LDL) cholesterol, LDL:HDL ratio], and biomarkers of stress and inflammation [cortisol, C-reactive protein, and interleukin 6 (IL6)]. Upon arrival to the laboratory and after resting seated for at least 5 min, vital signs (systolic and diastolic blood pressure, heart rate) were measured in duplicate via an automated digital sphygmomanometer (Omron M6 AC, Omron Inc., Kyoto, Japan) and averaged. Duplicate tests were conducted with a minimum of 2 min between tests. Blood pressure measurements were considered accurate if no more than 5 mmHg variation occurred between measurements. In addition to the above, participants were queried on adverse events (AE) throughout the trial.

To control for changes in exercise, sleep, and diet, participants were asked to keep exercise, sleep, and diet constant throughout the study period. To monitor adherence, exercise, sleep, and food logs were submitted throughout the study. Participants were required to write down the hours and type of exercise performed daily such as strength, cardio, or a combined strength/cardio exercise (CrossFit, yoga, etc.). The exercise logs were checked each visit to ensure that participants did not exercise 24 h prior or exercise strenuously 48 h prior to study visits. The logs were also evaluated to confirm participants did not exercise more than 6 h/week or less than 1 h/week, the limits set by the inclusion criteria to which participants agreed to keep constant throughout the trial. Hours of sleep per night were also recorded. Average weekly hours of exercise and sleep were evaluated at each study visit. A pre-visit food log was captured during the 24 h prior to baseline (Day 0), photocopied and returned to the participants so that they could repeat this diet the day prior to each visit. A 3-day food log was recorded for three non-consecutive days (two weekdays and one weekend day) between each visit, so any changes in diet could be calculated.

An evaluable sample size of 53 subjects per group was expected to provide 80% power (two-sided, α = 0.05) to detect an 8% difference in the number of hits in an assessment of reactive agility utilizing the Makoto device [44, 45]. A sample of at least 69 subjects per group was randomized to account for attrition and non-compliance (Fig. 1).

One hundred and forty-two individuals were randomized, 69 into the PLA and 73 into the PSE groups. Fifty-four and fifty-two participants completed the trial in the PLA and PSE treatment groups, respectively. Demographics are shown in Table 1. Treatment groups were well-balanced in demographics for the baseline characteristics. Although exercise as self-reported on the New Zealand Physical Activity score was different at the screening visit (p = 0.01), comparison of the actual hours of exercise collected on study diaries between screening and baseline visits (Day 0) indicated no differences between groups in exercise.

Compliance was not significantly different at any timepoint. At Day 90, compliance was 95.4 and 96.8% in the PLA and PSE groups, respectively, with no difference between groups (p = 0.328). Attrition among all participants was 25.4%, and it was not significantly different between treatments or genders (p = 0.5048).

Data and statistical analyses for outcomes obtained utilizing the Makoto device are shown in Table 2. The hits obtained with the stationary test with footplates for 30 s (Fig. 4) were significantly higher in PSE compared with PLA at Day 30 (PSE vs. PLA: 28.96 ± 2.08 vs. 28.09 ± 1.92 hits; p = 0.040; d = 0.435) and Day 90 (PSE vs. PLA: 28.42 ± 2.54 vs. 27.02 ± 3.55 hits; p = 0.002; d = 0.454). ART for the stationary test with footplates measured for 30 s were significantly lower in PSE compared with PLA at Day 7 (PSE vs. PLA: 0.5896 ± 0.060 vs. 0.6141 ± 0.073 s; p = 0.049; d = − 0.367) and Day 30 (PSE vs. PLA: 0.5811 ± 0.068 vs. 0.6033 ± 0.055 s; p = 0.049; d = − 0.359). The hits obtained with the multi-directional test with footplates (Fig. 5) for 30 s were significantly higher in PSE compared with PLA at Day 30 (PSE vs. PLA: 19.25 ± 1.84 vs. 18.45 ± 1.48 hits; p = 0.007; d = 0.479) and Day 90 (PSE vs. PLA: 19.39 ± 1.90 vs. 18.66 ± 1.64 hits; p = 0.026; d = 0.411). Significant differences were not observed in the remaining Makoto tests.

Plasma cortisol, IL6 and CRP were measured at all timepoints and are shown in Fig. 6. Comparisons between PSE and PLA were not significantly different for cortisol and IL6. Differences at baseline (p = 0.03) were present for CRP; therefore, ANCOVA was utilized to identify a significant treatment x visit interaction (p = 0.049) and pairwise comparisons revealed that PSE significantly decreased CRP at Day 30 compared to PLA (p = 0.045).

Adverse events were collected for all participants throughout the study. There was no difference in the total number of AE between groups (p = 0.5935). In addition, the total number of AE related to study product did not differ between groups (p = 0.3020). Complete blood count, complete metabolic profile, lipids, body metrics, and vital signs were assessed to evaluate overall health status over the study period. No overall treatment effects were observed for any of the outcomes (Additional file 1: Tables S1-S4). Four outcomes: absolute monocytes, absolute granulocytes, granulocyte percentages, and systolic blood pressure did evince significant treatment x visit interactions; however, any differences evident at day 30 were not significant at day 90 and all values remained within normal limits and were not deemed clinically relevant.

Evaluation of diet records collected at each study period indicated no difference between groups in total calorie, protein, fat, or carbohydrate intake (Table 3). Participants appeared to have maintained consistent diet over the 90-day study period and self-reported sleep and exercise also appeared to have been kept constant and did not differ between groups (Table 4).