No announcement yet.

An object lesson in getting what you want out of a study

  • Filter
  • Time
  • Show
Clear All
new posts

  • An object lesson in getting what you want out of a study

    Researchers had the idea of feeding cats a "moderate carbohydrate" diet to see what would happen to their gut microbiome. They found that cats fed on this diet had a higher level of saccharolytic bacteria and lower levels of proteolytic bacteria -- unsurprisingly.

    Here's where it gets interesting. "Live Science" reporting this headlines the story:

    High-protein diet not so good for kitty's belly, study suggest
    The study seems to have suggested nothing of the sort.

    Further, "Live Science" gushes:

    “Surprisingly, kittens eating the higher-carb kibble had more of the beneficial gut bacteria bifidobacterium. Low levels of bifidobacteria have been linked to irritable bowel syndrome in humans. Compared with the other groups of kitties, the high carb-eating kittens also had higher levels of lactobacillus, beneficial bacteria that may play a role in cholesterol and appetite regulation.”*

    Haven't located the study at the BMJ yet, but what you find elsewhere on the net is interesting and revealing.

  • #2
    Here we go:

    Twelve of the kittens became part of the study. The researchers took fecal samples at weaning and 4 and 8 weeks after weaning. They extracted bacterial DNA and used bioinformatics techniques to estimate total bacterial diversity.



    The kittens fed the MPMC diet had high levels of bifidobacteria, which was linked to higher blood ghrelin levels. Ghrelin is a hormone that stimulates appetite and thus may be linked to weight gain.


    Although kittens fed the HPLC diet had lower levels of some health-promoting bacteria, including Bifidobacterium, Lactobacillus, and Megasphaera, all the animals were healthy throughout the study.

    Kittens: Their microbiomes are what they eat

    So "Live Science" reported the link to leptin but elided the link that was noted to ghrelin.

    The study does seem to described some of the carb-digesting bacteria as "health promoting" (a dubious designationn I'd have thought: surely this depends on context) but did state that all cats on both diets seemed healthy.

    We also note all this is based on a sample of only 12 (sic) kittens and only what's supposedly deducible from only 3 (sic) fecal samples on which "bioinformatics techniques" which we're not told about have been used to get estimates.

    How in the world could "Live Science" deduce what it did from that? How would this be more telling, for example, than Pottenger's experiments that examined several variables, tracked physiological markers, and lasted for years?

    The experiment's interesting up to a point. But let's not try to get out of it what's not there -- and what runs counter to common sense. When did you last see any member of the cat family in the wild baking cup cakes?