5

votes

Can you help design experiment that pits Paleo Diet against USDA Food Pyramid?

Answered on August 19, 2014
Created July 12, 2011 at 9:17 PM

Which rat gets fat?

Introduction:

Background: The USDA recommends 45-65% carbohydrate, 10-35% protein and 20-35% fat calories for Adults(19 years and older).

-USDA Dietary Guidelines 2010 -pg. 15

The following statement suggests the USDA recommends eliminating saturated fat.

The body uses some saturated fatty acids for physiological and structural functions, but it makes more than enough to meet those needs. People therefore have no dietary requirement for saturated fatty acids.

The USDA seems to advocate replacing ALL saturated fat with PUFA:

replacing (saturated) with monounsaturated and/OR polyunsaturated fatty acids is associated with low blood cholesterol levels, and therefore a lower risk of cardiovascular disease.

-USDA Dietary Guidelines 2010 -pg 24

This experiment hypothesizes that when two rats are fed an equal number of calories, the rat that eats according to USDA Dietary Guidelines for Americans grows fatter than the rat that eats a high saturated fat diet designed by Paleo scientist (fill-in-the-blank).

Procedure:

Start with 2 rats, A + B.

Subject A Diet (adjusted for 2000 calorie diet):

Fat - 35% with 100% as polyunsaturated vegetable oil (Corn oil) Carbohydrate - 45% with 50% as whole grain wheat fiber and 50% as fruit/vegetables Protein - 20%

Subject B Diet (adjusted for 2000 calorie diet):

Fat - 35% with 90% saturated fat and 10% Omega 3 fats (Fish oil) Carbohydrate - 45% with 100% as fruits/vegetables/0 whole grains Protein - 20%

If I wanted to advance research that links Omega-6 fats and whole grain wheat fiber to obesity, how could I tweak this research project or modify this to pursue a Master's Thesis? Does the USDA rig their numbers so that it is impossible to generate obesity with their given ratios? How could I improve the design of this experiment?

Medium avatar

(10611)

on July 14, 2011
at 02:32 AM

Well when I got to Berkeley it was an eye-opening experience....I was doing funded research and the topics available were narrow. 18 months, a lot of experiments and writing later, I had a thesis. Bambam, if you're reading this, you're not going to get a master's degree for proving a foregone conclusion. What if your paleo rats pork out? You have to accept that and explain it, and be able to defend what you say.

Medium avatar

(12379)

on July 13, 2011
at 11:02 PM

hey thhq i wasn't the one that downvoted i think all your other points are great - but i think that # 3 was a little harsh - i know that the OP needs much more help on study design - but I think he came here for help

Medium avatar

(12379)

on July 13, 2011
at 11:02 PM

hey thhq i wasn't the one that downvoted i think all your other ponts are great - but i think that # 3 was a little harsh - i know that the OP needs much more help on study design - but I think he came here for help

Medium avatar

(10611)

on July 13, 2011
at 09:57 PM

Downvote me if you like, but I'm speaking the truth. I did an MS thesis at Berkeley. Funding research is tough and I think you'd be on your own on the above topic. It could be fleshed out, but doing a thesis at a major school using 4 rats isn't going to fly.

Medium avatar

(12379)

on July 13, 2011
at 04:18 PM

geez thhq - #3 is pretty rude don't you think?

71af94295988d55cd3b8340e619729d0

(255)

on July 13, 2011
at 10:56 AM

Why not have a 3 groups: a SAD, 'low-carb paleo', and a 'higher-carb paleo' group?

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 13, 2011
at 02:38 AM

And the statement is SUGGESTIVE of elimination of saturated fats because 1)Your body makes "more than enough" to meet needs and 2)People have "no dietary requirement" for saturated fats. How much less suggestive can they be without calling for us to outright eliminate saturated fats altogether?

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 13, 2011
at 02:35 AM

But for the record, they advice less than 10% of fat calories from saturated fat - the rest from vegetable oils. They recommend 10x more Omega 6 than 3. Am I supposed to give them credit for including trace amounts of Omega 3 when they have marginalized those fatty acids so harshly? Not easy for me.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 13, 2011
at 02:33 AM

Perhaps not but that statement looks like it could be interpreted by enough people to mean that - which pisses me off. It easily could've been written word for word by large food companies heavily leveraged in vegetable oils. I mean, it's extremely difficult for me to play nice when they fight dirty.

667f6c030b0245d71d8ef50c72b097dc

(15976)

on July 13, 2011
at 02:24 AM

I don't think it's fair to say that the government Suggests eliminating sat fat based on the quote you posted. They're merely reporting the fact. Like me saying you don't need carbohydrate - that does not imply that I feel you should not eat carbohydrate.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 13, 2011
at 02:07 AM

Yeah will good thing I have 3 years to work on this while I'm still an undergrad.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:59 PM

Well, I'm more interested in the effects of gluten grains like wheat on mammals and I would like to link it to obesity but that's difficult to do without manipulating fat type. They work together.

Medium avatar

(2169)

on July 12, 2011
at 11:56 PM

You should do a whole grain VS tuber carb experiment!! I can't believe I'm this excited about a lab experiment that doesn't involve anything exploding. Man, you're gonna need lots and lots of rats...

Medium avatar

(2169)

on July 12, 2011
at 11:54 PM

Once you discover how different fats alone impact their blood and body fat, then you can do another experiment that involves % of fat in the diet instead of type.

Medium avatar

(2169)

on July 12, 2011
at 11:53 PM

I think that's the way to go. And maybe pick the lowest GI carbs you can to limit triglyceride spikes? They need to eat the exact same things, all save the fats.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:51 PM

Ok, so all things fixed except fat type? That means I won't be feeding grains to either but all carbs from same source (fruits/veggies) right?

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:49 PM

Sweeeeet. I really want to punch the USDA in the teeth if I can or at least put out some research that proves while saturated fat goes up, weight goes down if amount of Omega6 and grains are controlled.

Medium avatar

(2169)

on July 12, 2011
at 11:49 PM

@BAMBAM- I think at "paleo" or "low carb" group is definitely in order! Although you might want to save that for a later experiment. Get the conclusions from fat type VS fat type and once you have a solid, reproducible conclusion, go to 50% carb vs. 70% fat. If you were to do that now, you've got way too many variables. try to design your experiment so that only one or two things are different between groups. Otherwise you won't be able to prove that one specific difference caused the difference between groups.

Medium avatar

(2169)

on July 12, 2011
at 11:44 PM

@BAMBAM- my best friend is currently getting her graduate degree in Public Health from Emory. She worked for the EPA on a big huge air quality study (where they actually had to document the relationship between air quality and sperm count, and yes, boxers or briefs was actually relevant!! *snort*) and did a bunch of research for the Department of Homeland Security on epidemics in farm animals in the US. I've sent her a link to this question with hope she'll have a little time to look it over for you. She'll know more than me about all the controls and stats and things.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:40 PM

What if I double the saturated fat in the Paleo rat group and feed them 70% calories as saturated fat?

Medium avatar

(2169)

on July 12, 2011
at 11:35 PM

Isn't a hypothesis an educated guess, and the experiment is conducted as a way to support or refute the hypothesis? I would say that his hypothesis is an educated guess. How could he phrase his hypothesis in a way that "supports" neither side? And I do agree with you, many scientists tweak experiments to turn out in favor of their hypothesis.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:32 PM

+1 from me. I consider your input invaluable. I hope to be able to combine all the contributions in this thread so that this whole community can design a kick ass experiment. There is a weight loss section of the USDA Dietary Guidelines but it is so general. They definitely are crafting statements so as to avoid being caught in a scientific pickle. They can make us fat without getting caught

Medium avatar

(2169)

on July 12, 2011
at 11:32 PM

that crossed my mind too. but because both sets of rats are getting equal ratios wouldn't that limit that phenomena? Or is it that saturated fat is more satiating than mono/poly and thus the rats fed saturated fat will inevitably not eat as much?

D3ff004d4a0c42b67cc2c49a5ee9c0f3

(5801)

on July 12, 2011
at 11:28 PM

Also, his hypothesis (rat grows fatter) is stating that he has already determined the conclusion before he begins his experiment. This is common and why sometimes scientists tweak the data.

D3ff004d4a0c42b67cc2c49a5ee9c0f3

(5801)

on July 12, 2011
at 11:26 PM

The problem with equal calories in these experiments is that fat and protein provide greater satiety and the subject voluntarily limits calories confounding if it was the limited calories or the diet.

Medium avatar

(12379)

on July 12, 2011
at 09:52 PM

hey BAMBAM - i think that stats got interesting for me when i finally understood what i would use them for - in school it was hard because the examples were all pretty useless to me - but now that i have large data sets i understand what the heck i learned all the stats for (and honestly I refer to my text book an awful lot!) Plus - my final year my stats prof was a sex psycologist - so his examples were AWESOME!!! Do the Nutrition Education MS - it looks like a great program - and like I said with the right context stats become interesting!

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 09:28 PM

+1 Bree from me. I'm horrible at stats and I'm eyeballing the Nutrition Education MS at Columbia University. Maybe I should reconsider. LOL.

  • 226b10cbb6b1d3530b00d2d84a2dc86e

    asked by

    (3313)
  • Views
    1.7K
  • Last Activity
    1426D AGO
Frontpage book

Get FREE instant access to our Paleo For Beginners Guide & 15 FREE Recipes!

4 Answers

5
Medium avatar

(12379)

on July 12, 2011
at 09:25 PM

You have a great question here - but in order to have a study that will have statistical merits - you will need more than 2 rats. At a minimum you will want 10 in each study group- as well you will want a 'control' group. I'm at a loss as to what you would feed the control group - but you want a group to compare both rat a and rat b to.

I would suggest dusting off the 'ole stats textbook and really diving into what kind of numbers you would need and also figuring out what tests that you will want to do on your data in order to 'prove' what you want.

This would be a very interesting study! And if you need help with stats - i suppose i could dust off my old text book and help!

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 09:28 PM

+1 Bree from me. I'm horrible at stats and I'm eyeballing the Nutrition Education MS at Columbia University. Maybe I should reconsider. LOL.

Medium avatar

(12379)

on July 12, 2011
at 09:52 PM

hey BAMBAM - i think that stats got interesting for me when i finally understood what i would use them for - in school it was hard because the examples were all pretty useless to me - but now that i have large data sets i understand what the heck i learned all the stats for (and honestly I refer to my text book an awful lot!) Plus - my final year my stats prof was a sex psycologist - so his examples were AWESOME!!! Do the Nutrition Education MS - it looks like a great program - and like I said with the right context stats become interesting!

1
Dbb1d3e11ab77b034fddac1def8d917e

on July 13, 2011
at 01:39 AM

Hey, I'm Ali's friend doing my MPH. I've got a few thoughts looking through this.

1- You've got WAY more than a master's thesis here. One of these experiments alone would be ambitious for a master's thesis.
2- What sort of funding and facilities do you have access to? You'll need to meet regulations and have IACUC approval for an experiment like this to be accepted by the scientific community. 3- Typically the level for statistical significance is around 30, depending on the power you're trying to achieve. That said, funding shortfalls make experiments with 5-10 rats/mice per group very common. 4- Why rats, specifically? Seems like the same thing could be accomplished for less expense or with more statistical power with the standard BALB or C57 strain mice. 5- With regards to groups, you have two options. You can either pick one thing to vary between two groups (plus a control group on standard lab chow) or you can have several groups and more dietary variations. With the latter option, you may be limited in what you can claim by comparing the groups to each other, hence the importance of a control group. 6- Be very very careful assuming the results of your experiments while you're designing them. Go to the library or find a professor with an expertise in scientific ethics and ask for help. Statements like "I really want to punch the USDA in the teeth" are going to sink your credibility, and with it, the results of your research, if they ever get out. Look at ClimateGate. And also, one study (or even a series of studies) is never going to form the basis of policy change. The agency that tried it would get torn apart by lobbyists and the courts. If you want to change policy with scientific studies, you'll need to get a group of researchers to duplicate your work and expand upon it until you have a body of evidence from a variety of people that sends a clear message. At the EPA, we started to hit that tipping point when we had 15-20 studies showing similar results. The USDA is much more traditional and under the sway of the food industry, so you'll need way more than that. 7- In terms of experimental design, I'd probably start with something like this. 1, control group eating standard lab chow. 2, group eating USDA standard diet. 3, group eating low-carb paleo diet. 4, group eating higher-carb paleo diet. That should give you enough to do some nice analysis, and give you questions to answer for your next experiment. ;) 8- And speaking of analysis, start thinking now about the tests you'll need to do to get the information you want. You'll have to document all of it for IACUC approval, and most likely for funding approval as well. Regular blood draws sound like they'd be in order, as well as body composition estimates and autopsies at the end of the experiment. How long do you want to keep this running? Does the information you want require special sacrifice methods? At this point you're getting past my expertise, but I know that you'll need to think about it.

I've written enough of a novel now, so I'll stop here. Hope this was helpful!

1
Medium avatar

on July 12, 2011
at 11:16 PM

BAMBAM- Sweet idea. Are you going to seriously go for it? Good luck and I hope you find out some cool stuff. I can't help you on what to feed the control rats, but here are a few things that I thought of:

  1. You definitely need 10+ rats for A and 10+ B. If you can get more than 10 go for more. More is always better especially when you're using statistics.

  2. You say "the rat that eats according to USDA Dietary Guidelines for Americans grows fatter" but I think this needs to be more specific. The USDA is making these recommendations with reference to blood cholesterol levels and thus cardiovascular disease, not obesity. We make assumptions that these are related. You can simply replace the "gets fatter" with something along the lines of "increases dangerous small LDL blood cholesterol particles" Or you could go all out and hypothesize that the USDA rat will not only have bad blood levels but will also become obese/gain body fat.

  3. Which gets me to my next point. What about blood sugar? Fructose is sugar. Shouldn't the paleo rats be fed low GI tubers? Aren't high triglycerides linked to cardiovascular disease?

  4. Also, are you going to call rat B the "paleo" rat? Because I think "paleo" definitely isn't necessarily that high carb, but I guess it isn't necessarily "low carb" either, it just depends on the individual needs of the dieter? How many paleo eaters get 1/2 their calories from carbs? Don't high carbs cause high triglycerides = cardiovascular disease? (Obviously you probably know more than me in this area, which is why all these statements have ? marks at the ends)

PS. my degree is in outdoor recreation (yes I have a degree in FUN). I took a few research methods and stats classes. This is definitely not where my expertise lies but I figured I would share my 2 cents.

Medium avatar

(2169)

on July 12, 2011
at 11:35 PM

Isn't a hypothesis an educated guess, and the experiment is conducted as a way to support or refute the hypothesis? I would say that his hypothesis is an educated guess. How could he phrase his hypothesis in a way that "supports" neither side? And I do agree with you, many scientists tweak experiments to turn out in favor of their hypothesis.

D3ff004d4a0c42b67cc2c49a5ee9c0f3

(5801)

on July 12, 2011
at 11:28 PM

Also, his hypothesis (rat grows fatter) is stating that he has already determined the conclusion before he begins his experiment. This is common and why sometimes scientists tweak the data.

Medium avatar

(2169)

on July 12, 2011
at 11:53 PM

I think that's the way to go. And maybe pick the lowest GI carbs you can to limit triglyceride spikes? They need to eat the exact same things, all save the fats.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:40 PM

What if I double the saturated fat in the Paleo rat group and feed them 70% calories as saturated fat?

D3ff004d4a0c42b67cc2c49a5ee9c0f3

(5801)

on July 12, 2011
at 11:26 PM

The problem with equal calories in these experiments is that fat and protein provide greater satiety and the subject voluntarily limits calories confounding if it was the limited calories or the diet.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:32 PM

+1 from me. I consider your input invaluable. I hope to be able to combine all the contributions in this thread so that this whole community can design a kick ass experiment. There is a weight loss section of the USDA Dietary Guidelines but it is so general. They definitely are crafting statements so as to avoid being caught in a scientific pickle. They can make us fat without getting caught

Medium avatar

(2169)

on July 12, 2011
at 11:49 PM

@BAMBAM- I think at "paleo" or "low carb" group is definitely in order! Although you might want to save that for a later experiment. Get the conclusions from fat type VS fat type and once you have a solid, reproducible conclusion, go to 50% carb vs. 70% fat. If you were to do that now, you've got way too many variables. try to design your experiment so that only one or two things are different between groups. Otherwise you won't be able to prove that one specific difference caused the difference between groups.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:59 PM

Well, I'm more interested in the effects of gluten grains like wheat on mammals and I would like to link it to obesity but that's difficult to do without manipulating fat type. They work together.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:49 PM

Sweeeeet. I really want to punch the USDA in the teeth if I can or at least put out some research that proves while saturated fat goes up, weight goes down if amount of Omega6 and grains are controlled.

Medium avatar

(2169)

on July 12, 2011
at 11:44 PM

@BAMBAM- my best friend is currently getting her graduate degree in Public Health from Emory. She worked for the EPA on a big huge air quality study (where they actually had to document the relationship between air quality and sperm count, and yes, boxers or briefs was actually relevant!! *snort*) and did a bunch of research for the Department of Homeland Security on epidemics in farm animals in the US. I've sent her a link to this question with hope she'll have a little time to look it over for you. She'll know more than me about all the controls and stats and things.

Medium avatar

(2169)

on July 12, 2011
at 11:54 PM

Once you discover how different fats alone impact their blood and body fat, then you can do another experiment that involves % of fat in the diet instead of type.

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 12, 2011
at 11:51 PM

Ok, so all things fixed except fat type? That means I won't be feeding grains to either but all carbs from same source (fruits/veggies) right?

Medium avatar

(2169)

on July 12, 2011
at 11:56 PM

You should do a whole grain VS tuber carb experiment!! I can't believe I'm this excited about a lab experiment that doesn't involve anything exploding. Man, you're gonna need lots and lots of rats...

Medium avatar

(2169)

on July 12, 2011
at 11:32 PM

that crossed my mind too. but because both sets of rats are getting equal ratios wouldn't that limit that phenomena? Or is it that saturated fat is more satiating than mono/poly and thus the rats fed saturated fat will inevitably not eat as much?

71af94295988d55cd3b8340e619729d0

(255)

on July 13, 2011
at 10:56 AM

Why not have a 3 groups: a SAD, 'low-carb paleo', and a 'higher-carb paleo' group?

-2
Medium avatar

(10611)

on July 13, 2011
at 02:03 AM

We have some problems here Houston

  1. Not enough rats.
  2. Rat studies prove rat theories.
  3. Forget about Berkeley or Stanford. Pray that Ball State takes you.
  4. You have to normalize the diets for equal taste. If you use tasty cookies for group A you have to use smoked dry rub ribs for group B. Or gruel and boiled chicken, your choice.

Aside from all that why is the paleo diet so high in carbs?

Medium avatar

(12379)

on July 13, 2011
at 04:18 PM

geez thhq - #3 is pretty rude don't you think?

Medium avatar

(12379)

on July 13, 2011
at 11:02 PM

hey thhq i wasn't the one that downvoted i think all your other points are great - but i think that # 3 was a little harsh - i know that the OP needs much more help on study design - but I think he came here for help

Medium avatar

(10611)

on July 13, 2011
at 09:57 PM

Downvote me if you like, but I'm speaking the truth. I did an MS thesis at Berkeley. Funding research is tough and I think you'd be on your own on the above topic. It could be fleshed out, but doing a thesis at a major school using 4 rats isn't going to fly.

Medium avatar

(12379)

on July 13, 2011
at 11:02 PM

hey thhq i wasn't the one that downvoted i think all your other ponts are great - but i think that # 3 was a little harsh - i know that the OP needs much more help on study design - but I think he came here for help

226b10cbb6b1d3530b00d2d84a2dc86e

(3313)

on July 13, 2011
at 02:07 AM

Yeah will good thing I have 3 years to work on this while I'm still an undergrad.

Medium avatar

(10611)

on July 14, 2011
at 02:32 AM

Well when I got to Berkeley it was an eye-opening experience....I was doing funded research and the topics available were narrow. 18 months, a lot of experiments and writing later, I had a thesis. Bambam, if you're reading this, you're not going to get a master's degree for proving a foregone conclusion. What if your paleo rats pork out? You have to accept that and explain it, and be able to defend what you say.

Answer Question


Get FREE instant access to our
Paleo For Beginners Guide & 15 FREE Recipes!