As a counterbalance to the papers documenting the obesogenic effect of PUFA containing diets I'd like to have a brief aside about the Surwit diet. There's a nice list of the ingredients here.
The first thing to say is that it is based around hydrogenated coconut oil. I would guess this is fully hydrogenated coconut oil, though the manufacturers do not specify. To supply a small amount of PUFA some soybean oil is added to give just under 7% of fat as PUFA which, at 60% of calories as fat, translates to 4% of total calories as PUFA. That is low yet the diet is one of the most obesogenic on the market.
There is more to obesity than PUFA (gasp).
The whole drive of the Surwit diet appears to be to supply glucose in a form which will be rapidly absorbed and so allow penetration beyond the liver and in to the systemic circulation. Maltodextrin is used for this. The diet uses a medium chain length polymer of glucose which has a glycaemic index in the same ball park as pure glucose.
Systemic hyperglycaemia is combined with sucrose at 17% of calories to induce pathological insulin resistance and a little extra hyperglycaemia. This is combined with the normal physiological insulin resistance associated with oxidising saturated fats. The need for extremely high levels of insulin to maintain normoglycaemia appears to be adequate to force lipid storage in the face of adipocyte insulin resistance. It certainly works, even with minimal PUFA.
Then I got side tracked in to Surwit's early work. A great paper is this one using F1850 for a publication in 1988:
Diet-Induced Type II Diabetes in C57BL/6J Mice
F1850 is still interesting despite having a bit more linoleic acid (around 7% of calories). This still feels like quite a low dose for a mouse diet. The paper has some great graphs. Let's have a look and ask some questions.
Take two mouse strains (Bl/6J and A/J mice), feed both on sucrose, maltodextrin and modest linoleic acid lard supplying about 7% of calories as PUFA. Both strains get fat, like this:
The bottom two lines are the two strains of mice on standard crapinabag and we can ignore these. Today's quiz is to identify the top two lines based on the fasting blood glucose/insulin data.
Here are the fasting insulin and glucose levels after 24 weeks of exposure to either control or obesogenic diet:
Grey columns represents the crapinabag fed mice, black columns the obesogenic F1850 diet fed mice. It is quite clear that while the A/J mice became modestly insulin resistant it was the Bl/6J mice which became severely insulin resistant, with marked increases in fasting insulin and glucose. On the same diet as the A/J mice.
Now, match the mouse strain to the upper traces on the weight gain graphs. We can ignore the crapinabag fed lower lines. Of the two upper lines one represents a set of mice which is fatter than the other. Are B/J or Bl/6J fattest? From the insulin data...
If you carry the idea in your head that insulin causes obesity you might expect the hyperinsulinaemic Bl6/J mice with a fasting insulin of 150microU/ml to be fattest. If you consider that the mice on sugar become fat because they find the sugary food more "rewarding" (it's the same food, idiot!) you might come to the idea that obesity is caused by overeating and that obesity secondarily causes insulin resistance (no sniggering at the back there). And both are incorrect.
The fattest mice are the A/J mice. They are fattest because they are the most insulin sensitive.
Insulin signalling, not insulin per se, causes fat storage.
The Bl/6 mice, on the same diet as the B/J mice, are more insulin resistant so resist insulin induced weight gain. Significant weight gain still happens because gross hyperinsulinaemia can overcome insulin resistance. But the greater weight gain for the A/J mice is much easier by using less insulin combined with lower insulin resistance.
The A/J mice become "fat but fit" based on glucose/insulin levels. Given long enough they will, undoubtedly, join the Bl/6J mice but they have a longer way to go to become adequately obese to resist insulin. But I have no doubt this will happen on F1850.
I think the difference between the strains is genetic and that Bl/6 based mice generate enough ROS to generate significant insulin resistance at much lower levels of insulin signalling than B/J mice do. If I had to guess at a gene I would go for their truncated supercomplex assembly protein causing failure of supercomplex formation and so allowing electron leakage to molecular oxygen from the ETC at times when no such thing should happen, at least not until the CoQ couple is a great deal more reduced.
But that's just speculation, which is fun.
Humans look more like B/J mice (normal ETC) than Bl/6J mice to me.
Bottom line: Insulin signalling is what makes you fat. Insulin resistance resists this. It's all in the data.
Addendum. Insulin sensitivity is a Good Thing. But it makes you fat, which is a Bad Thing. Let's rephrase that: Insulin sensitivity is a Good Thing provided you don't use it! Perhaps that concept is worth a post in it's own right.