The Chronicle of the Horse
MagazineNewsHorse SportsHorse CareCOTH StoreVoicesThe Chronicle UntackedDirectoriesMarketplaceDates & Results
 
Results 1 to 17 of 17
  1. #1
    Join Date
    Nov. 13, 2010
    Posts
    2,236

    Default SmartGut Ultra...with clinical research!

    SmartGut Ultra
    Saw smartpak's latest supplement, smartgut ultra. Anyone think it will be the "next big thing"? I have to say I'm not impressed by their "research".



  2. #2
    Join Date
    Feb. 1, 2012
    Location
    Vermont
    Posts
    5,262

    Default

    Count me in as another who was not impressed with the "research"...

    n=8...come on! They couldn't find more than 8 horses!
    And, if I'm reading the graph correctly, the control group and the research group had the exact same number of ulcers so how again does the SmartGut Ultra results in less ulcers for the horses taking it, versus the ones who aren't?!
    "If you think nobody cares about you, try missing a couple payments..."



  3. #3
    Join Date
    Feb. 1, 2012
    Location
    Vermont
    Posts
    5,262

    Default

    Here is the link to the "research":

    http://www.smartpakequine.com/pdfs/S...ch_Summary.pdf
    "If you think nobody cares about you, try missing a couple payments..."



  4. #4
    Join Date
    Jan. 16, 2002
    Location
    West Coast of Michigan
    Posts
    36,321

    Default

    How you can get p values like that from 8 horses escapes me . . .

    What *is* striking is the obvious rebound recurrence of ulcers. Supports the idea that PPIs
    should be weaned more than it supports the addition of nostrums and potions, IMO.
    Click here before you buy.


    2 members found this post helpful.

  5. #5
    Join Date
    Feb. 1, 2012
    Location
    Vermont
    Posts
    5,262

    Default

    Delta - are you reading the graph the same way that I am? That at the end, the horses fed SGU had the same number as the control group, thereby showing that SmartGutUltra doesn't mean less ulcers?
    "If you think nobody cares about you, try missing a couple payments..."



  6. #6
    Join Date
    Jul. 10, 2008
    Location
    NC
    Posts
    951

    Default

    Quote Originally Posted by SuckerForHorses View Post
    Count me in as another who was not impressed with the "research"...

    n=8...come on! They couldn't find more than 8 horses!
    And, if I'm reading the graph correctly, the control group and the research group had the exact same number of ulcers so how again does the SmartGut Ultra results in less ulcers for the horses taking it, versus the ones who aren't?!

    The P value determines statistical significance; doesn't really matter the n= as long as it's over 3 really.

    http://frank.mtsu.edu/~dwalsh/436/CORRSIG.pdf
    Quote Originally Posted by rustbreeches View Post
    [George Morris] doesn't always drink beer, but when he does, he prefers Dos Equis


    1 members found this post helpful.

  7. #7
    Join Date
    Jul. 10, 2008
    Location
    NC
    Posts
    951

    Default

    Quote Originally Posted by deltawave View Post
    How you can get p values like that from 8 horses escapes me . . .

    What *is* striking is the obvious rebound recurrence of ulcers. Supports the idea that PPIs
    should be weaned more than it supports the addition of nostrums and potions, IMO.
    I do agree with this.

    ETA---To me, this study doesn't look like SmartGutULTRA does anything to prevent or help with the recurrence of ulcers longterm. It does do something a bit ~14 days later (after the cessation of omeprazole) I would say, but it seems to have some sort of tapering off effect later on.
    Quote Originally Posted by rustbreeches View Post
    [George Morris] doesn't always drink beer, but when he does, he prefers Dos Equis


    1 members found this post helpful.

  8. #8
    Join Date
    Feb. 1, 2012
    Location
    Vermont
    Posts
    5,262

    Default

    Quote Originally Posted by jlphilli View Post
    The P value determines statistical significance

    http://frank.mtsu.edu/~dwalsh/436/CORRSIG.pdf
    Be definition, yes. I think DW's point is that how can you say there is statistical significance with only 8 test subjects? That's a pretty small test group.
    "If you think nobody cares about you, try missing a couple payments..."



  9. #9
    Join Date
    Jul. 10, 2008
    Location
    NC
    Posts
    951

    Default

    Quote Originally Posted by SuckerForHorses View Post
    Be definition, yes. I think DW's point is that how can you say there is statistical significance with only 8 test subjects? That's a pretty small test group.
    I actually don't think that's a small test group for a large animal study. Like I said...as long as it's more than 3, the P value determines significance. It just doesn't matter (statistics).
    Quote Originally Posted by rustbreeches View Post
    [George Morris] doesn't always drink beer, but when he does, he prefers Dos Equis


    1 members found this post helpful.

  10. #10
    Join Date
    Feb. 1, 2012
    Location
    Vermont
    Posts
    5,262

    Default

    Quote Originally Posted by jlphilli View Post
    I actually don't think that's a small test group for a large animal study.


    Really???
    "If you think nobody cares about you, try missing a couple payments..."



  11. #11
    Join Date
    Jul. 10, 2008
    Location
    NC
    Posts
    951

    Default

    Quote Originally Posted by SuckerForHorses View Post


    Really???
    Considering this isn't a Nature paper, yes.
    Quote Originally Posted by rustbreeches View Post
    [George Morris] doesn't always drink beer, but when he does, he prefers Dos Equis


    1 members found this post helpful.

  12. #12
    Join Date
    Jul. 10, 2008
    Location
    NC
    Posts
    951

    Default

    Here's some examples of "legit" studies:

    http://www.ncbi.nlm.nih.gov/pubmed/23186182 (5 horses)
    http://www.ncbi.nlm.nih.gov/pubmed/23206314 (9 horses)
    http://www.ncbi.nlm.nih.gov/pubmed/23206274 (13 horses)
    http://www.ncbi.nlm.nih.gov/pubmed/23206252 (7 umbilical cord samples)
    http://www.ncbi.nlm.nih.gov/pubmed/23205506 (62 horses using medical records only)

    ETA--I didn't pick and choose; I went from the top 5 of the Equine Vet Journal
    Quote Originally Posted by rustbreeches View Post
    [George Morris] doesn't always drink beer, but when he does, he prefers Dos Equis


    2 members found this post helpful.

  13. #13
    Join Date
    Mar. 24, 2004
    Location
    Central PA
    Posts
    1,269

    Default

    According to this report, there were 8 horses per treatment. This is actually a pretty good number of animals for a study with horses, especially one that requires repeated gastroscopy and measurements of gastric pH. In other studies done by this group, they have used a total of 8 horses in a crossover design. Each horse serves as its own control, and this is a good statistical tool when limited numbers of animals are available.

    This report doesn't reveal what was done between days 35 and 42, so we can't determine why the ulcer score there decreased. I suspect the horses received another course of GastroGard during that time. I was unable to find this information published in a peer-reviewed journal.


    2 members found this post helpful.

  14. #14
    Join Date
    Jan. 16, 2002
    Location
    West Coast of Michigan
    Posts
    36,321

    Default

    No, a n=8 is actually fairly typical for this type of study in the veterinary world. So while I would still call it "small" one could also argue that it is the norm. Especially for what is probably a non-funded study.

    And I'm no statistician, but to get very significant p values (<0.05) with just a few subjects requires that the endpoints being looked at vary quite widely, no? I do hope someone more statistically fluent than I can clarify, but the results shown do not appear to me to match up with anything that would give such a low p value. Seems more to me like torturing the data to make things look "significant".

    It is an effort, and any company trying to make an effort gets kudos from me. I just think the data presented says a lot more about the nature of ulcers and the perils of stopping omeprazole abruptly (although ulcers did seem to respond well in these 8 horses in only 14 days) than it does about the wonders of their supplement.
    Click here before you buy.



  15. #15
    Join Date
    Mar. 24, 2004
    Location
    Central PA
    Posts
    1,269

    Default

    You do not need wildly different endpoints to get P < 0.05 with 16 horses. If they did, indeed, use a crossover design, this would reduce the contribution of animal-to-animal variation even more. It is entirely plausible to have P < 0.05 in this situation. That said, there is not enough information in this report to adequately evaluate the research. With this number of animals, they could show the raw data rather than ulcer scores. A horse with an ulcer score of 1 might have 2 lesions, while a horse with a score of 2 might have 3 lesions. One could argue that this is not a meaningful difference. However, with this number of animals, sometimes it's necessary to put the responses into categories to glean useful information. This is one reason why an individual study should not be viewed as the "answer". As a scientist, it's also important to consider the possibility that the product does have an effect. You have to be skeptical, but fairly evaluate the information. That said, I would not recommend this product based on the information they presented.


    2 members found this post helpful.

  16. #16
    Join Date
    Jul. 10, 2008
    Location
    NC
    Posts
    951

    Default

    Quote Originally Posted by Skip's Rider View Post
    You do not need wildly different endpoints to get P < 0.05 with 16 horses. If they did, indeed, use a crossover design, this would reduce the contribution of animal-to-animal variation even more. It is entirely plausible to have P < 0.05 in this situation. That said, there is not enough information in this report to adequately evaluate the research. With this number of animals, they could show the raw data rather than ulcer scores. A horse with an ulcer score of 1 might have 2 lesions, while a horse with a score of 2 might have 3 lesions. One could argue that this is not a meaningful difference. However, with this number of animals, sometimes it's necessary to put the responses into categories to glean useful information. This is one reason why an individual study should not be viewed as the "answer". As a scientist, it's also important to consider the possibility that the product does have an effect. You have to be skeptical, but fairly evaluate the information. That said, I would not recommend this product based on the information they presented.
    ^
    Quote Originally Posted by rustbreeches View Post
    [George Morris] doesn't always drink beer, but when he does, he prefers Dos Equis



  17. #17
    Join Date
    Jan. 16, 2002
    Location
    West Coast of Michigan
    Posts
    36,321

    Default

    Thank you for that!
    Click here before you buy.



Similar Threads

  1. Clinical Research/CRAs?
    By mango1612 in forum Off Topic
    Replies: 1
    Last Post: Nov. 25, 2011, 09:46 PM
  2. Clinical research coordinators out there?
    By jen-s in forum Off Topic
    Replies: 3
    Last Post: Sep. 11, 2011, 11:00 PM
  3. EPM - clinical signs - how did YOUR horse present?
    By Meadow36 in forum Horse Care
    Replies: 12
    Last Post: May. 10, 2011, 02:24 PM
  4. Let's play clinical case study (please?)
    By champagnetaste in forum Horse Care
    Replies: 17
    Last Post: Mar. 20, 2010, 01:29 AM
  5. Anyone using SmartDigest Ultra or SmartGut??
    By FoxChaser in forum Horse Care
    Replies: 40
    Last Post: Mar. 17, 2009, 12:05 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
randomness