|
| Does Garlic Help To Improve VO2max? (Read 546 times) |
|
|
posted: 3/30/2008 at 8:35 PM
modified: 3/30/2008 at 8:40 PM |
Interesting. Even though the results weren't substantially different, they were statistically signifcant. Pretty harmless. Why not try it? Kwai garlic is readily available in U.S.
http://www.vitacost.com/KwaiEveryDay#IngredientFacts |
| Those who try, fail! Those who do what it takes to succeed, succeed!! |
|
|
|
posted: 3/30/2008 at 11:39 PM |
I can almost guarantee that if you swallow a heap of garlic before a run, most people aren't gonna be wanting to run anywhere near your breath!!! So you'll either be waaaaay in front, or everyone will be passing you so they don't have be running downwind  |
|
|
|
posted: 3/31/2008 at 12:53 AM |
| Ha! That is why you take the odorless garlic in supplement form. |
| Those who try, fail! Those who do what it takes to succeed, succeed!! |
|
|
|
posted: 3/31/2008 at 12:58 AM |
| Yeah, but it wouldn't be as much fun as watching runners dropping away from you now would it!!! |
|
|
| view log Ms. Fussypants |
posted: 3/31/2008 at 1:11 AM |
| You'd think the Italians would be putting out more great athletes. |
|
|
| view log Funky Monkey |
posted: 3/31/2008 at 9:37 AM |
There is no way to know based on the information provided in the article.
In an RCT, the characteristics of each group, while randomly assigned, may still differ. For example, the garlic group may be younger or more likely to be male. All good RCTs compare demographics between the groups to prove that randomization truly equalized these potentially confounding variables. This garlic study did not.
Does garlic help to improve VO2 Max? Maybe. Maybe not. But this study provides NO reliable evidence either way. |
| It's all fun and games until the flying monkeys attack. |
|
|
| view log Funky Monkey |
posted: 3/31/2008 at 9:47 AM |
Also, on re-reviewing the results, this study had NOWHERE near the power to achieve the P values they got. They state that they used T testing, which is reasonable, but with so few subjects and confidence intervals that overlap between groups, I am suspicious that the P values reported are incorrect. Since they provide the actual data, I am tempted to repeat the analysis.
But first, I'm off to run. Which is more likely to improve performance...
 |
| It's all fun and games until the flying monkeys attack. |
|
|
| view log Trample the Weak |
posted: 3/31/2008 at 2:06 PM |
Quote from Trent on 3/31/2008 at 9:47 AM:But first, I'm off to run. Which is more likely to improve performance...
You made me read this. I hope you're happy. WTF were you doing up at 4:37?
|
E-mail: JakeKnight2002@aol.com
"The past is nothing but a series of recollections; it does not own you ... if we are prisoners of the past, we are jailer as well."
~~ Jack Kerley, The Hundredth Man
Unedited Running-Related Song Lyrics
You're a runner
but you're chasing yourself
Feel a hot breath on your shoulder.
Oh, wait. That's just Tanya.
~~ DIO
Less thinking, more running. |
|
|
| view log Funky Monkey |
posted: 3/31/2008 at 4:00 PM |
Log stalker.
Garlic gives me gas anyway.
|
| It's all fun and games until the flying monkeys attack. |
|
|
| view log Trample the Weak |
posted: 3/31/2008 at 4:58 PM |
| Quote from Trent on 3/31/2008 at 4:00 PM:
Garlic gives me gas anyway.
I'm not sure that can get worse. |
E-mail: JakeKnight2002@aol.com
"The past is nothing but a series of recollections; it does not own you ... if we are prisoners of the past, we are jailer as well."
~~ Jack Kerley, The Hundredth Man
Unedited Running-Related Song Lyrics
You're a runner
but you're chasing yourself
Feel a hot breath on your shoulder.
Oh, wait. That's just Tanya.
~~ DIO
Less thinking, more running. |
|
|
| view log Playmaker / nemesis |
posted: 3/31/2008 at 5:46 PM
modified: 3/31/2008 at 5:47 PM |
| Quote from Trent on 3/31/2008 at 9:47 AM: Also, on re-reviewing the results, this study had NOWHERE near the power to achieve the P values they got. They state that they used T testing, which is reasonable, but with so few subjects and confidence intervals that overlap between groups, I am suspicious that the P values reported are incorrect. Since they provide the actual data, I am tempted to repeat the analysis.
I hate myself for doing this, but I repeated their analysis. For a paired t-test, I obtained p-value of 0.0051. The reason this value is so surprisingly low is that it is a paired t-test as opposed to a standard t-test. Using a standard t-test for means assuming equal variances, I got p=0.4554 (two-tailed). However, since these are not two random groups, but rather, the same individuals being measured twice, the paired test is indicated, and the researchers chose correctly.
However, I'm still dubious of the conclusion because of the sample size. And something doesn't smell right (excuse the pun). I can barely believe that not one of the subjects had a higher VO2max with the placebo. I mean, you're doing the same test two weeks apart, but since it's a treadmill test rather than actual VO2max measured from breath analysis, you think at least one person would have felt kind of crappy on their random "garlic" day and not been able to stay on the treadmill as long.
Also seems odd when you look at the delta between garlic and placebo
# garlic placebo delta
1 48.8 48.8 0
2 63.5 60.6 2.9
3 60.6 57.6 3
4 54.7 51.8 2.9
5 51.8 51.8 0
6 63.5 60.6 2.9
7 63.5 63.5 0
8 57.6 54.7 2.9
9 57.6 54.7 2.9
10 51.8 51.8 0
So garlic either has no effect, or increases your VO2max by 2.9 (or 3.0) in one case.  Seems fishy to me.
If I continue to hate myself this afternoon, I may do more rigorous analysis with our DOE software in the lab (as opposed to Excel, which I used). Speaking of software, who the f**k needs SPSS to do a friggin' t-test?
|
20th Century: 800m: 2:04 |1600m: 4:37 |3200m: 10:06 |5k: 16:23 |10k: 35:38 |15k: 54:20 |25k: 1:35:59
21st Century: 5k: 19:42 |10k: 43:00
What are you doing? |
|
|
| view log |
posted: 3/31/2008 at 5:50 PM |
| Paging Dr. Dilligaf. Dr. Dilligaf, please report to the garlic lab. |
|
|
| view log Funky Monkey |
posted: 3/31/2008 at 5:53 PM
modified: 3/31/2008 at 5:55 PM |
Were the data normally distributed?
The confidence intervals overlap. The outcome cannot be significant. What is the rationale for pairing? Just because they are the same subjects with and without the intervention does not necessarily justify this unless they can demonstrate that all other factors are equal. And they cannot be since there may have been a learning effect.
And I stand by my concern that they did not look for confounders. |
| It's all fun and games until the flying monkeys attack. |
|
|
| view log Playmaker / nemesis |
posted: 3/31/2008 at 6:17 PM
modified: 3/31/2008 at 6:25 PM |
| Quote from Trent on 3/31/2008 at 5:53 PM: Were the data normally distributed?
n=10 is too small to really tell if the data are normally distributed, so that's a red flag right there.
Quote from Trent on 3/31/2008 at 5:53 PM:The confidence intervals overlap. The outcome cannot be significant.
True, but in a paired test, you are actually testing whether the mean difference is statistically different from zero. Surprisingly, this is significant in their study despite 4/10 subjects having a difference of zero. This is because the differences were not normally distributed.
MTA: the mean difference is 1.75 with s=1.506. The 95% confidence interval for the mean difference comes out to 1.75±0.93. [±1.96*sqrt(s/n) 1.96 is the point on the standard normal curve corresponding to alpha=0.05 (95% of the data) ] Since this interval does not contain zero, the difference in means is statistically significant.
Quote from Trent on 3/31/2008 at 5:53 PM:What is the rationale for pairing? Just because they are the same subjects with and without the intervention does not necessarily justify this unless they can demonstrate that all other factors are equal. And they cannot be since there may have been a learning effect.
And I stand by my concern that they did not look for confounders.
Their rationale for pairing is that they're the same subjects before and after treatment with garlic. They justify this by saying the garlic application was random (some test garlic, then placebo, others the opposite). I agree with you, Trent; that's not enough control.
Too small of a study -- 4/10 showed no difference, sample size too small, pair assumption is dubious.
I love me some garlic, but maybe I should just run more. |
20th Century: 800m: 2:04 |1600m: 4:37 |3200m: 10:06 |5k: 16:23 |10k: 35:38 |15k: 54:20 |25k: 1:35:59
21st Century: 5k: 19:42 |10k: 43:00
What are you doing? |
|
|
|
|
