Nutrition supplementation with antioxidants have been discussed as a way to further enhance overall well-being of athletes, promote faster recovery, and improve overall performance. The use of Protandim, a nutritional supplement containing 5 botanicals (Bacopa extract 150 mg; milk thistle 225mg; ashwagandha 150 mg; green tea 75 mg; turmeric 75 mg) has shown promise in an earlier study by reducing oxidative stress and increasing the amount of the anti-oxidant enzymes in the blood.

The purpose of this study was to examine the effect of ~90 day Protandim supplementation on 5-km running performance and on acute and long term oxidative damage as assessed by blood markers. Secondarily, another purpose of this study examined the effect of Protandim supplementation on other blood parameters (such as antioxidant enzyme concentrations) and measures of quality of life. The experiment was double-blind, placebo controlled. The study addressed two main questions and two secondary questions:

Main Question 1: Does regular supplementation of Protandim (675 mg/day for 90 days) improve 5-km running times? Hypothesis: Protandim would improve 5-km running time by 0.5 min (SD 1-min). This is equivalent to an effect size of 0.5. The placebo group will have no change in performance.

Main Question 2: Does regular supplementation of Protandim (675 mg/day for 90 days) reduce oxidative damage at rest as assessed by lipid peroxides (TBARS) in runners? Hypothesis: Based on the data by Nelson et al. (2006), oxidative damaged would be reduced by 40% (effect size = 4.8).

Secondary Question 1: Does regular supplementation of Protandim (675 mg/day for 90 days) in runners reduce the increase in oxidative damage post-race compared to pre-race as assessed by lipid peroxides (TBARS)? Hypothesis: Based on the data by Kyparos et al. (2009), we expected a 45% increase in TBARS post-exercise compared to pre-exercise (effect size of 3.6). However, we expected that those that are on Protandim for 90 days will only have 31% increase in TBARS post-race (effect size = 2.5) [based on changes in pre-post lipid peroxides from Arent et al. (2010)].

Secondary Question 2: Does regular supplementation of Protandim (675 mg/day for 90 days) improve quality of life according to WHOQOL-BREF (Skevington, Lotfy, O'Connell, & Group, 2004)? Hypothesis: There would be a statistical improvement in the quality of life post-Protandim supplementation which a small effect size improvement of 0.33.

Forty apparently healthy community runners 18-55 yrs of age was recruited to participate in this study that will require six total visits to the University of Louisville, which included three sessions at the running track at University of Louisville.

A about 120 mL of blood was withdrawn from a peripheral vein for the duration of the study (about 105 days). Runners was recruited from running clubs across the local community. These participants will be randomized into two groups (Protandim, Placebo). The groups will be sex and performance matched after the initial 5-km race.

Each subject underwent six testing sessions.

Session 1 (Initial Screening Day)

Subjects signed consent form and completed a questionnaire to clear them for physical activity. Age, weight, and height was recorded. Subjects was asked to list all the nutritional supplements they are taking, including multivitamins. They were asked to list the 5-km personal best time, and when that was done. They reported most recent 5-km time and when that was achieved. Then they were asked to refrain from taking any multivitamins or nutritional supplements for the duration of the study. The subjects recorded what they ate during the 24 hr prior to the test day, and then they were asked to eat the same thing and do same activity before the all follow up tests. Then a training diary was given to them to fill in for the duration of the study. The training diary included intensity, training duration, and mileage per week. After, they were given six quality of life questionnaires [World Health Organization's WHOQOL-BREF quality of life assessment to fill out at home (one each for Session 1-6)]. A pregnancy test was also conducted in the females.

Session 2 (Baseline, About 15 days after Session 1)

Subjects were fasted in the morning when they arrived at University of Louisville for their pre-exercise blood sampling. They brought their 24-hr dietary recall with them and their quality of life forms. Approximately three teaspoons of blood was withdrawn 30-60 minutes prior to exercise for analysis of several blood parameters. Then subjects were required to partake in a 5-km time trial at the University of Louisville track. The runners raced together in groups of 6 to 10 runners per race according to their estimated 5-km time. The estimated 5-km time is their own best estimate. Time trials have greater logical validity compared to time-to-exhaustion tests so that is why time-trials are selected as the performance measure.

About 10-20 minutes post-exercise, another three teaspoons of blood was withdrawn from each subject. Only after their post-exercise blood draw will the subjects be able to eat.

Based on the performance of the 5-km time trials, males and females were ranked separately. This was done in randomized blocks of two for males, and then randomized into blocks of two for females. For example, the males was ranked from the 5-km race from the fastest runner to the slowest runner. First and second place males were placed in a block of two. Third and fourth place males were placed in a block of two. And so on. Then a fair coin was flipped. If heads, the first ranked runner in the first pair went to Group 1, otherwise, the second ranked runner in the pair went to Group 1. This was done for each block of two males (each pair of two males). The same is then performed for the women. This randomization allowed for the same mean 5-km running times in each group as well as maintaining similar male to female ratios.

Session 3 (About 7 days after Session 2)

Subjects were fasted in the morning when arrived at University of Louisville for their pre-exercise blood sampling. They brought their 24-hr dietary recall with them and their quality of life forms. Approximately three teaspoons of blood were withdrawn 30-60 minutes prior to exercise for analysis of several blood parameters (the analytes are listed in Table 3). Then subjects partook in a 5-km time trial at the University of Louisville track.

About 30-minutes post-exercise, another three teaspoons of blood were withdrawn from each subject. Only after their post-exercise blood draw were the subjects be able to eat.

Depending on the group, they were given approximately a 100 day supply of Protandim or Placebo (corn starch and food coloring). The study was double blinded so neither the researchers nor the subjects knew which group is the Protandim group and which group was the Placebo group. They ingested 1 pill per day, ideally with breakfast (675 mg per day, 1 pill per day). The subjects were given the complete supply of pills during this session, right after the second blood sample was taken. The subjects was classified as Group 1 or Group 2 based on their five kilometer results from the first session. The pill bottles were labeled Group 1 or Group 2. The pill bottles were dispensed by the research team according to group number after their second blood sample. Subjects were asked record the daily intake of the pills in their training diary. Once the study was completed, subjects were asked to return the pill bottles to the research staff. That way, the researchers were able to account for any missed days.

Session 4 (About 23 Days after Session 3)

Subjects were fasted in the morning when arrived at University of Louisville for their pre-exercise blood sampling. They brought their 24-hr dietary recall with them and their quality of life forms. Approximately three teaspoons of blood were withdrawn. There was no 5-km running race performed in this session. A pregnancy test was performed in the females.

Session 5 (about 30 Days after Session 4)

Subjects were fasted in the morning when they arrived at University of Louisville for their pre-exercise blood sampling. They brought their 24-hr dietary recall with them and their quality of life forms. Approximately three teaspoons of blood were withdrawn. There was no running race performed in this session. A pregnancy test was conducted in the females.

Session 6 (About 30 days after Session 5)

Subjects were fasted in the morning when they arrived at University of Louisville for their pre-exercise blood sampling. They brought their 24-hr dietary recall with them and their quality of life forms. They also brought any unused pills for proper documentation. A pregnancy test was conducted in the females. Then, approximately three teaspoons of blood were withdrawn 30-60 minutes prior to exercise for analysis of several blood parameters. Then subjects partook in a 5-km time trial at the University of Louisville track.

About 30-minutes post-exercise, another three teaspoons of blood was withdrawn from each subject. Only after their post-exercise blood draw were the subjects be able to eat.

Things to Consider: The between subject-variability in oxidative stress biomarkers is large (Mullins et al. 2013) and suggests that there may be responders and non-responders to an oxidative stress challenge (Mullins et al. 2013). Individuals who have large increases in oxidative stress (i.e. responders) could be reflecting an inability to regulate redox homeostasis and warrant additional support (i.e. with antioxidant supplementation with Protandim). It is also necessary to obtain week-to week and month to month coefficient of variation in markers of oxidative stress so that meaningful changes in these markers can be evaluated. Thus, the placebo group allowed us to obtain the week-to-week and month to month coefficient of variation. As well, for the same reason, it is necessary to obtain week-to-week and month to month variation in running performance, thus, the placebo group will allow for the development of the coefficient of variation in performance.

Measuring dietary data is troublesome. First, questionnaires don't give details about antioxidant intake - rather they give outcomes like fruit and veg intake, etc. More in depth measures wouldn't be worth the trouble. Thus, having subjects record what they eat (and their physical activity) during the 24 hr prior to the test day, and then having to ask them to eat the same thing and do same activity before the follow up tests at seven days and 90 days was required.

Environmental conditions was recorded on the days of the three time trials (wind, temperature, % humidity, barometric pressure).

Statistical Analyses

Once all the data was collected and analyzed, LifeVantage revealed which group was which. The main dependent variables measured were the 5-km finishing time, TBARS, SOD, and both the quality of life questionnaire. Based on a 5-km time improvement of 2.5% or about 30 seconds (SD = 1 minute) with Protandim, and no improvement in the placebo group, about 32 runners in total would be needed (Effect size = 0.5, statistical power = 80%, alpha error probability = 5%, Correlation amongst repeated measures = 0.90, F-test Family, ANOVA repeated measures, between factors, G*Power 3.1.2, Universität Kiel, Germany). Accounting for a 20% attrition rate (8 subjects), a total of 40 subjects were recruited (20 per group).

Possible Risks of Ingesting Protandim: We did expect there to be any side effects for the typical Protandim consumer. However, according to the Protandim website (http://www.protandim.com/faqs), some individuals have natural allergic responses to one of the ingredients, just as some people are allergic to pine pollen or penicillin. These allergic responses to Protandim generally appear as gastrointestinal disturbances (i.e., stomach ache, diarrhea, vomiting) or sometimes as a headache or rash on the hands or feet. The symptoms disappear if Protandim is discontinued. However, according to the Natural Standard Database (https://naturalmedicines.therapeuticresearch.com/), each botanical has been shown to have rare side effects. They are listed there.

Inclusion criteria:

1. Healthy, non-pregnant subjects between 18 and 55 years of age. Apparently healthy is classified as A-1, A-2, or A-3 according to the American Heart Association

   http://circ.ahajournals.org/content/97/22/2283/T6.expansion.html

2. Subjects have to achieve a classification of "local class" based on age-graded time USA Track & Field. The age-graded time is the finish time adjusted to that of an open division participant using a factor for age and gender. Thus, the times for women and older participants are adjusted downward, while the times for most open division participants (such as 25-year-old men) remain the same. For example, a 55 year old woman has to run 29 minutes 45 seconds or better to be in the study:

   http://www.usatf.org/statistics/calculators/agegrading/

   For a man the same age, the equivalent time would be 25 minutes 2 seconds. This is approximately 60% of the speed of the current world record time for that age.

3. Subjects who will abstain from taking any nutritional supplements for the duration of the study, including vitamins and mineral supplements (Exception, ferrous sulfate, elemental iron, Vitamin D, Calcium). Subjects will also abstain from taking any over the counter products (herbals, melatonin, St. John's Wort, etc…) for the duration of the study.

The exclusion criteria will be the following:

1. Under 18 and over 55 years of age; those who are not apparently healthy is classified as A-1, A-2, or A-3 according to the American Heart Association.
2. Subjects who are not able to run 5-km in the time required for their age and gender.
3. Subjects that will continue to take nutritional supplements, including over the counter products, for the duration of the study, including vitamins and mineral supplements (exception: Ferrous sulfate, Elemental iron, Vitamin D, Calcium).
4. Subjects that are taking prescription medications with the exception of birth control.
5. Known allergy or sensitivity to milk thistle, Bacopa monnieri, Ashwagandha, turmeric (or ginger), tea, its parts, caffeine, tannins, or members of the Theaceae family.

The de-identified data is in SPSS format placed as a supplement in the published PLOS ONE article: http://dx.doi.org/10.1371/journal.pone.0160559

Over a period of 3 months (88 days)

• Serum TBARS to detect lipid peroxidation is not specific.
• TBARS may have been measured too quickly post-exercise.