It was bound to happen

This blog entry is courtesy of Fran Mayo who read an article in “Runner’s World” about some of the benefits of drinking pomegranate juice. It was fairly inevitable that I would get around to talking about pomegranate juice. It is, after all, all the rage right now and POM is currently in the media regarding some dubious health claims.

There are lots of reasons to drink pomegranate juice. Personally, when it first came out as a commercial product, I thought it was a pure novelty. I mean, have you ever EATEN a pomegranate? It takes FOREVER. The whole idea of juicing enough fruit to make a whole bottle of pure pomegranate juice was just unfathomable. So from my perspective, one of the reasons to drink pomegranate juice is because you can. All that pomegranate-y taste without the painstaking work.

But apparently, there are other reasons to drink pomegranate juice apart from the pure sensory ones. And apparently, one of them is to possibly decrease muscle soreness. Or so this study claims. Fran did all the work to track it down (would it be so much work for a magazine to actually cite a study instead of saying, “A study from Joe Schmoe University says…”)

Trombold JR, Barnes JN, Critchley L, et al. Ellagitannin consumption improves strength recovering 2-3d after eccentric exercise. Medicine and Science in Sport and Exercise. 42(3) 493-498, 2010.


There is a theory that chemicals called polyphenols are good for you. The ways in which these benefits are quantified are variable across studies and the practical significance of these findings may or may not be actually that large, but that’s not where I’m going with this review.

The authors of this study thought that delayed-onset muscle soreness would be a good model to study the effects of polyphenols on some aspect of functional improvement (as opposed to physiological improvements which may or may not result in functional improvements). There is good evidence to show that eccentric exercise does cause local inflammation and other physiological changes which could be considered similar, if only transitory, to more serious types of chronic inflammation as are observed in diseases such as rheumatoid arthritis.

The authors reviewed some of the literature on eccentric exercise recovery and other interventions. In their literature review, they could not find any supplements that were able to demonstrate improved muscle function recovery by reducing inflammation or oxidative stress.

The purpose of this study therefore, was to figure out whether pomegranate juice improves muscle function over a four-day recovery period after eccentric exercise.


This study was a double-blind (a long-archaic term now) randomized placebo controlled crossover experiment. This means that all of the subjects received both interventions (placebo and real pomegranate juice); that two parties (usually the subject and the investigator) were blinded to which subjects got what drink; and that the subjects were randomly assigned with respect to which intervention they got first.

Potential subjects were considered for the study if they were non-smokers, and “recreationally active”. However, potential subjects were excluded from the study if they had done any resistance training in the past 3 months, if they were in a formal endurance training program, or had a history of previous upper extremity injury. Potential subjects were also excluded if they had had a “recent” weight change of >5kg, hypertension, or if they were on anti-inflammatory medications or other similar drugs.

Subjects were assigned to receive either a placebo or pomegranate drink in two testing periods, each one 9 days long with a 14 day washout period. During the testing period, subjects drank two 480ml bottles of drink each day, 12 hours apart. On day 5 of the testing period, the subjects then performed an eccentric bicep curl exercise (2 sets of 20 reps, with one rep every 15 seconds, and an eccentric phase of 3 seconds and a rest period of 4 minutes between sets.) The curls were done on a Cybex machine. The subjects used a randomly selected arm for the first test period and then the other arm for the second test period.

Strength was measured at 2 hours, and then 1, 2 and 3 days after eccentric curls and was measured isometrically with a rigged up modified preacher curl and load cell. Testing was performed only on the exercised arm.

Soreness was measured using a visual analog scale. Blood tests were also performed for creatine kinase, myoglobin, interleukin-6 and C-reative protein.

Statistics: The authors used repeated measures ANOVAs and the least significant differences method for pair-wise comparisons to compare the effects of the drinks as well as the outcome values over time. Six ANOVAs were performed, one for each outcome variable.

[The method in which the subjects were randomly assigned to their drink order was not reported. The method in which the arm was randomly chosen for each test period was not reported.]


Sixteen males were recruited for the study. Average age was 24 ish years.

When it came to soreness, or any of the blood tests, the authors failed to detect a statistically significant difference between the placebo and pomegranate juice. When we look at the values between the two drinks, there don’t appear to be any differences that I would consider large enough to warrant a discussion about power.

The main result of this paper is the detection of a single difference between pomegranate juice and the placebo in recovery strength. In both trials, a decrease in isometric strength was observed (an almost 30% decrease in strength in both conditions). At 2 hours, and 24 hours, there was no difference detected between the two drinks in strength. At the 48 hour mark, the authors detected a statistically significant difference in strength, with pomegranate juice being associated with higher strength than the placebo. This difference was also statistically significant at 72 hours (day 3). By day 4, this difference had disappeared.

At the 48 hour period, the difference in percent of baseline strength was 7.1%. Strength in the pomegranate condition was, on average 85.4% of baseline compared to 78.3% in the placebo; both with standard deviations of 10.1%. The difference at 72 hours was 4.9%, with strength being 88.9% for pomegranate and 84% for placebo (SD 7.9 ish for both conditions).

[However, the ACTUAL value of the baseline strength was never published, so we are unable to determine the actual value this nearly 30% decrease represents and therefore can’t really interpret how significant this decrease is with respect to something practical like, for instance, how much less weight might this translate to?]

The authors did report confidence intervals around some of their “strength deficit” values.

At 48 hours, the mean recovery in strength for the pomegranate condition was 13.9% with a 95%CI of 7.84 to 19.9%.
At 48 hours, the mean recovery in strength for the placebo condition was 5.5% with a 95%CI of 7.84 to 2.38 to 8.68%.

[This should set off a small flag in the back of your head if you’re familiar with confidence intervals, as these intervals overlap, which calls the significant p-value into question as possibly spurious. It’s not a perfect rule of thumb, but not a bad one as far as rules of thumbs go.]

There was also a significant “ordering effect” detected and reported, where strength was also dependent on the trial (whether it was the first testing period or second testing period). That is to say that strength values were influenced by whether they were collected in the first or second trial, in favor of the second trial. This effect was noted at the 2 hour, 72 hour and 96 hour marks. Nine of the subjects started with placebo drinks, while 6 subjects started with the pomegranate drinks. [Oddly, there were supposed to be 16 subjects in this study as reported above. I don’t know what happened to that last guy.]


I think this study is one of the few studies in which the major leak springs up in the introduction of the manuscript and then is followed by subsequent smaller, but contributory leaks. The authors of this study theorized that since DOMS takes several days to recover from, that strength can be used as a reliable indicator of the time course of functional recovery.

I think this theory is wrong. And here’s why:

DOMS is primarily manifested by pain, and while DOMS can be so severe so as to keep you from wanting to move or lift, I don’t believe the presence/absence/severity/duration of DOMS can be considered an accurate reflection of any reliable physiological parameters, never mind a functional one such as strength. In fact, the loss of strength that is associated with DOMS may be more pain limited than it is physiologically limited such that if you could simply work through the additional pain, you might not have any strength decrements at all 24-48 hours after an initial eccentric exercise bout (which is the typical lag time between exercise and DOMS).

So given the imperfect model of how this is all going to happen, pomegranate juice comes out as having improved strength recovery, and the study is otherwise conducted in a fairly rigorous manner (with the standard reporting problems). However, we are still left with the confounding factors of pain limited recovery vs. physiologically limited recovery. This is re-enforced by the finding that there did not seem to be any differences in the inflammatory markers they selected to study between the placebo and pomegranate juice, which then begs the question, “Did ‘reduction of oxidative stress or inflammation’ actually occur as a result of the pomegranate juice?”

This is also one of those studies in which the use of ‘untrained healthy males’ probably does matter. If you’re an athlete or even an occasional “weekend warrior” (as the authors suggest as a population to which this study might be applicable), your ability to recover from exercise is not a constant and changes as your body enacts adaptations to the training you do. So, if we subscribe to the idea that recovery from activity gets better with training, this study (in the best case scenario) gives us an idea of what the MAXIMUM benefit would be that we could ever derive from pomegranate juice. Frankly, it seems rather underwhelming if I have to pay about 30 bucks for a bottle that has 16 servings in it (so basically an 8 day supply).

From a training perspective, I think this study has limited utility. There are very few activities that require repetitive eccentric loading to the point of incurring DOMS on a consistent basis (i.e. progressively increasing loads) that individuals participate in on such a frequency basis so as to have to worry about recovery within 4 days. Yes, skiers undergo cyclical eccentric loading, but unless they’re getting progressively heavier or moving to progressively steeper/higher frequency loading courses on a workout-to-workout basis in a constantly increasing scale (whether linear or curvilinear), the need to marginally accelerate recovery seems misplaced.

Furthermore to the use of the study for training reasons (which I really don’t think the authors actually set out to show), is the whole issue of whether post-exercise inflammation is something we actually want to prevent. If we subscribe to the theory that training causes tissue damage directly and indirectly, and performance/aesthetic improvements occur because these tissues heal/adapt such that they have higher work capacity/size, then inflammation is the first and necessary step of healing. Are we, in fact, doing harm to our fitness goals by preventing what is a physiologically necessary step (i.e. would our results as a result of blocking inflammation be as large as simply allowing the normal time course of healing to occur?) As an example, the trade-off question would be, “Would my biceps get even bigger if I could work them out twice a week at the same intensity because I partially blocked the inflammatory phase of tissue healing than if I just let things happen and trained either less frequently or at a lower TOTAL intensity (since the second bout would presumably be less than the first)?” Or is this a case of slower progression because the “build back stronger” effect is attenuated by blocking the inflammatory response? (i.e. I can lift the same amount of weight 3 days after, which enables me to train biceps twice a week, but the amount of net tissue accretion is actually close to 0 because the inflammatory response provides some sort of “build more than before” signal?)

So while there might not be any direct adverse/side effects from lowering things like “oxidative stress”, we are quite in the dark with regards to possible either longer-term indirect “harms” (i.e. not getting where you want to go as efficiently as possible) which might manifest as a result of chronic exposure to “inflammatory reducers”.

These are questions to which I do not believe we have the answers to. And so, tinkering around with training protocols as a result of single-exposure pre-post studies is, from the get-go, not a well-informed action.

All of this discussion is assuming that the “significant difference” is actually REAL. The authors never published the results of their ANOVAs (the test statistic, the degrees of freedom, the p-value), only the “protected” t-tests (as defined by the least significant difference method). With 6 ANOVAs and no adjustment, it’s entirely possible these “significant” p-values are, in fact, products of pure chance as opposed to reflections on real effects. The fact that there are overlapping confidence intervals in their data suggests that this possibility is not just theoretical.

But even if it IS real, it’s really hard to gauge the usefulness of this study. We are given no absolute values with regards to strength. A decrease of 30% in maximum strength could mean something as small as 10 pounds, or as great as 30 (I don’t know how strong “recreational, non-resistance trained” 24 year old males generally are). Therefore the fact that pomegranate juice seemed to be linked to a faster recovery in strength has no real meaning given that we don’t know what the actual magnitude of recovery was. Do I care that I can lift 4 extra pounds a day earlier? Would it matter if that was 20 extra pounds? And lastly, would I be willing to buy enough pomegranate juice (at approximately 2 dollars per drink) to drink it twice a day EVERY DAY to get that effect?

Click Here to view the Full Version of our Website