Category Archives: pair-feeding

Circadian Disruption Impairs Survival in the Wild

…just read that huge disasters, ranging from Exxon Valdez to Chernobyl, may have been due, in part, to ignorance of basic principles of circadian rhythms.  Gravitas.


circadian rhythms

Continue reading


Liver is evil but need not be punished. SFAs.

What to serve with a liquid lunch, and a recipe for chocolate.

It’s like a feed forward downward spiral.  If you don’t eat saturated fat & MCTs prior to imbibing, then liver intentionally makes more PUFAs for the alcohol-induced burning ROS to molest.  Liver is evil but need not be punished.  SFAs.

Brief background: (Kirpich et al., 2011 & 2013)

Researchers studying alcohol in rodents know where they’re going and like to get there fast.  70 drinks per day fast.  Granted, rats metabolize faster than humans so it’s likely a little less… but a little less than 70 is still a lot of sauce.

Continue reading


The curious effects of calories in mice

What is the biological impact of a history of obesity and weight loss?  The metabolic trajectory of two calorically restricted skinny mice depends entirely upon whether or not they used to be fat.  The end of this story might be: ‘Tis better to have lost and re-gained than never to have lost at all; or it’s just an interesting new take on the body weight set point theory.

Caloric restriction chronically impairs metabolic programming in mice (Kirchner et al., 2012)

divide and conquer

Part 1.
Study 1. Calorie restricted lean mice: the effect of diet composition.

Continue reading


Insulin per se

This recent manuscript nearly slid beneath the radar… almost stopped reading at the abstract until the word “nifedipine” appeared (among its widely pleiotropic effects, nifedipine also lowers insulin).

The series of experiments described below demonstrate one aspect of the scientific method reasonably well.  None of the individual experiments, when viewed in isolation, really prove the hypothesis.  But the researchers tested it with a variety of widely different methods and all of the results went in the same direction.  The hypothesis in question: insulin causes fat gain, and hyperinsulinemia per se, not macronutrients or calories, is the root cause.

This group has previously shown that sucrose is more detrimental than fish oil is beneficial toward obesity and glycemic control.

High glycemic index carbohydrates abrogate the anti-obesity effect of fish oil in mice (Hao et al., 2012 AJP)

Divide and conquer
Mouse study.  Lots of diets, in brief:
Pair fed: high fish oil (180 g/kg) plus 13%, 23%, 33%, and 43% sucrose (by weight, switched out for casein [a poor choice IMO])
High fish oil (180 g/kg) plus sucrose, fructose, glucose, low GI carbs, and high GI carbs.
That’s a lot of diets.  Kudos.

As expected, higher sugar and lower protein intakes enhance weight gain (yes, even when pair-fed similar calories [i.e., a calorie is not a calorie]) and this is at least partly due to reduced metabolic rate (as per the poor man’s energy expenditure test- measuring body weight before and after 24 hours starvation [higher weight loss = higher metabolic rate]):High sucrose-fed mice also had more inflamed adipose tissue and less thermogenic brown fat, which likely contributed to their glycemic dysregulation and elevated adiposity.

Sucrose is comprised of glucose and fructose, so to determine which component was causing obesity, they fed mice high fish oil diets plus either sucrose, glucose, or fructose.  Interestingly, the glucose group gained as much weight as the sucrose group.  Since the fructose group gained the least amount of weight, the researchers attributed the sucrose-induced obesity to insulin! (fructose doesn’t elicit an insulin response; and insulin levels were lowest in the fructose group).

Body weight, plasma insulin, and glucose tolerance:

I. Thus far: glucose and sucrose cause obesity by stimulating insulin secretion.  Glycemic deterioration is worst in the glucose-fed group because they were consuming most of the most insulinogenic sugar: glucose.  It was lower in the sucrose and fructose groups because sucrose contains only half as much glucose as pure glucose, and fructose contains no glucose.  IOW, these data suggest hyperinsulinemia per se causes obesity and insulin resistance.  Gravitas.

They further tested this by comparing high and low GI diets which cause higher and lower insulin levels, respectively.  As expected, the low GI diet led to less weight gain, and significantly lower insulin levels and adipose tissue accumulation compared to the high GI diet:

II. Thus far: high insulin levels, whether induced by glucose, sucrose, or high GI starch, lead to obesity.

They next took a non-dietary approach by artificially increasing insulin levels with glybenclamide in fish oil-fed mice to see if hyperinsulinemia could still cause obesity.  The results weren’t robust, but the higher insulin levels tended to increase adiposity even in mice fed the anti-obesogenic fish oil diet. 

In the experiment, the opposite approach was taken: nifedipine was used to lower insulin.  The use of octreotide and diazoxide has been used in a similar context with similar results in humans, discussed HERE and HERE.Again, the results were not robust, but when viewed collectively a picture begins to emerge: raising insulin levels, whether it is with a high glucose or sucrose diet, a high GI diet, or glybenclamide increases adipose tissue growth; and conversely, lowering insulin levels, whether it is with a less insulinogenic sugar diet (fructose), a low GI diet, or nifedipine decreases adipose tissue growth.  Oh yeah, and low carb works too.

calories proper


another side of leptin

Op. 56

Leptin is probably just as important as insulin WRT obesity, and this is as just as good a place as any to learn about this increasingly interesting hormone.

ABCs of Leptin in a nutshell:

A. Fed state: leptin is secreted from adipose and tells the brain to maintain food intake and energy expenditure at a body weight set point, which is likely established by diet.

B. Fasted state: leptin secretion declines, causing hunger to go up and energy expenditure to go down.

In the past, the cause and consequence of leptin resistance received a lot of my attention due to their importance in obesity.  Leptin resistance is, in brief, obesity.  Or the mouse on the right:

C. Obesity: eating a poor diet causes leptin resistance, which allows the body weight set point to rise until your fat cells stop responding to insulin (it’s kind of complicated)

But there’s another side of leptin that is mostly unknown, frequently overlooked, and poorly understood.  And I say this “is as just as good a place as any” to learn about it because while this side of leptin isn’t as popular as the energy expenditure, appetite, etc., stuff, it could very well be just as important, IMHO.

Leptin vs. the pathological hyperglycemia in diabetic state(s) (note the plural form of “state[s]”).

Leptin deficiency causes insulin resistance induced by uncontrolled diabetes (German, Morton, et al., 2011 Diabetes)

Divide and conquer

STZ is a beta-cell toxin used to induce diabetes.  STZ-treated mice have low insulin, low leptin, and lose weight despite a voracious appetite (just like type 1 diabetic humans).  Their insulin resistance is fully corrected while their marked hyperglycemia is attenuated by leptin injections.  Leptin reduces food intake but this doesn’t reduce body weight because energy expenditure paradoxically declines (discussed below).  Furthermore, diabetic mice restricted to eat only as much as leptin-treated diabetic mice (STZ-veh-PF)  lose significantly more weight because they lack the leptin-induced suppression in energy expenditure.

Summary of energy balance:

Control mice (veh-veh) eat the least but have much lower energy expenditure, causing them to weigh the most.  Energy expenditure is the more important variable driving high body weight in these animals (it goes down significantly more than food intake).  Diabetic mice (STZ-veh) eat the most food which is balanced by high energy expenditure (explained below), causing an intermediate body weight.  When the voracious appetite of diabetic mice is restrained (STZ-veh-PF), they weigh the least (they are starving).  Food intake is the more important variable driving low body weight in diabetic mice (energy expenditure is the same in diabetic and diabetic-PF mice).  STZ-leptin mice have intermediate food intake, energy expenditure, and body weight.  All is well, leptin cures the deranged energy balance of type I diabetes.

As mentioned above, diabetic mice eat more but weigh less because of drastically increased energy expenditure.  Energy expenditure is increased, in part, due to out-of-control gluconeogenesis (from hyperglucagonemia).  The paradoxical effects of leptin on energy expenditure (increases it in post-obese subjects but decreases it in diabetic mice) may be explained by leptin-induced reduction of this out-of-control gluconeogenesis, mediated via normalization of glucagon.

The authors further demonstrated that leptin restores liver, but not muscle or adipose insulin sensitivity in diabetic mice, independent of food intake.

Thus, insulin-deficiency -> dec. leptin -> inc. glucagon -> inc. hepatic glucose output -> hyperglycemia

Insulin’s primarily role might be to suppress glucagon.  STZ-induced “relative” state of starvation causes leptin to plummet; in the basal state, leptin may not have anything to do with glucagon because insulin keeps it under control.  But in diabetes, there’s no insulin to suppress glucagon; this is where exogenous leptin struts its stuff.

Collectively, these data further support the conclusion that insulin’s major function is to suppress glucagon, as opposed to other effects in skeletal muscle or adipose.  It’s been almost a year since Unger’s notorious publication which showed that glucagon receptor knockout mice were immune to type 1 diabetes (discussed here).  Diabetic hyperglycemia is partially mediated by insulin resistance, and largely mediated by hyperglucagonemia.  But why wasn’t hyperglycemia completely normalized by leptin replacement therapy?   These researchers sought simply to normalize leptin levels, but what would’ve happened if they provided a supraphysiological dose of leptin?  During physiological leptin replacement therapy, glucagon levels were normalized, but a hint of hyperglycemia remained.  Some other [leptin resistant?] member(s) of glucagon’s nefarious cohort must be responsible for the residual diabetic hyperglycemia…

Fortunately for us, the effects of supraphysiological leptin were tested in an identical experimental paradigm:

Leptin therapy reverses hyperglycemia in mice with streptozotocin-induced diabetes, independent of hepatic leptin signaling (Denroche, Kieffer, et al., 2011 Diabetes)

Indeed, supraphysiological leptin therapy overcame whatever diabetic “leptin resistance” remained and totally cured hyperglycemia.

This study repeated much of what was done in the first study, but whereas the first study added that leptin’s effect on food intake was not involved, this study showed that hepatic leptin signaling was not responsible either:

In both cases, hyperglucagonemia appears to be a major cause of diabetic hyperglycemia, and this is cured by leptin.  Insulin sensitivity was only partially restored by physiological leptin replacement; this seems to be due to some sort of apparent “leptin resistance,” which is overcome by supraphysiological leptin.  Diabetic insulin resistance is most likely caused by hyperglucagonemia-induced increased hepatic glucose output, and this is cured by leptin’s [non-hepatic] effects on reducing glucagon levels.  Diabetic insulin resistance may be partially caused by a brain mechanism, but at least one brain mechanism (food intake) was ruled out by the German study.

And oh so interestingly, all of these effects were mimicked by leptin administration directly into the brain, at a dose which caused no change in peripheral leptin levels (German, Morton, et al., 2011 Endocrinology).  The “STZ-lep” in the figure below refers to diabetic mice with leptin administered directly into the brain.

Back to the plural form of diabetic “state[s]” mentioned in the intro.  All of the above studies were in insulin-deficient SKINNY type 1 diabetic mice.  The next study is in OBESE type 2 diabetic rats.

Subcutaneous administration of leptin normalizes fasting plasma glucose in obese type 2 diabetic UCD-T2DM rats (Cummings, Havel, et al., 2011 PNAS)  

N.B. the control rats in this study were pair-fed to the leptin treated animals to control for the leptin-induced satiation.  Leptin-treated mice lost less weight most likely because energy expenditure declined (just like in the first study mentioned above).

In agreement with the glucose normalization seen by leptin treatment in skinny type 1 diabetic animals, leptin reduced glucose levels in obese type 2 diabetic rats, and this too was associated with reduced glucagon levels:

In type 1 diabetic animals, there are very low insulin levels regardless of food intake, leptin treatment, and hepatic leptin signaling.  Thus, insulin has nothing to do with the effects of leptin in type 1 diabetes.  This study showed that insulin levels also have nothing to do with the effects of leptin in obese type 2 diabetes:

So while the efficacy of exogenous leptin administration in established obesity is questionable, it is capable of combating the pathological glucagon-induced hyperglycemia which is responsible for much of the damage incurred by the diabetic state[s].

Leptin, glucagon, and diabetes.


calories proper


Trans fats, take III

Man vs. ape     Or
Postmenopausal women vs. Africa green monkeys


Two good and potentially somewhat contradicting studies on our good old friends, trans fats.

Trans fat diet induces abdominal obesity and changes in insulin sensitivity in monkeys. (Kavanagh et al., 2007 Obesity)

In this study, Kavanaugh and colleagues fed either a control diet or one fortified with trans fatty acids to a group of African green monkeys for 6 years.  Two immediate strengths of this study are 1) the use of primates, who respond to dietary intervention much more similarly to humans than rodents, and 2) the duration is long enough to model what would be seen in a human population.  Furthermore, to prevent differences in food intake from affecting the outcome, all of the animals were fed 70 kcal/kg of their initial body weight.  This feeding regimen was chosen specifically to prevent an energy imbalance, i.e., the monkeys were to be “weight stable” for the entire study.  This method is superior to pair feeding, where one group is fed ad libitum and the other group is given the same amount of calories as the first group, but instead of grazing all day (normal behavior) they get it all in one sitting.  Pair feeding is stressful for the animals and causes a whole host of other problems.  Both groups in this study received exactly the same amount of food (70 kcal/kg of initial body weight) every day for 6 years.

At the beginning of the study, the monkeys weighed ~6.5 kg (14.3 pounds); thus, for the rest of the study they were fed 455 kilocalories every day.  The diet consisted of 35% fat, 17% protein, and 48% carbohydrates.

The diet for half of the monkeys was supplemented with 8%, or ~4 grams, of trans fatty acids.  The average intake for humans is 3%, or ~7 grams per day.  An intake level of 8% for humans is around 18 grams, which could be accomplished by eating fast food or microwave popcorn a few times per week.  So besides being informative and shedding a new light on energy balance, this study is also of practical relevance.

Furthermore, the trans fat they chose was similar to the most abundant trans fat found in human diets (processed foods): partially hydrogenated soybean oil.

The diet:


TRANS refers to trans fatty acids, and CIS is the opposite of trans.  Cis fatty acids are the form of most natural fatty acids.  People don’t usually call regular fatty acids “cis” because it is assumed; this is how most unsaturated fatty acids are found in nature.

To the data.


divide and conquer.


Body weight was roughly similar in CIS (closed circles) and TRANS (open circles) but started to diverge toward the end of the study.

The control group (CIS) weighed 6.41 kg at baseline and 6.55 kg at follow-up, an increase of less than 2%.  This was expected because at baseline, 70 kcal/kg per day was precisely enough food to keep them weight stable, so essentially nothing changed in these monkeys.  More specifically, since food intake and body weight didn’t change, we can say that there were probably no major perturbations in energy balance in this group.


TRANS, on the other hand, gained almost 3 times more weight despite eating exactly the same amount of food as the control group, which was exactly the same amount of food they were eating when they were weight stable at baseline.  Energy balance was clearly perturbed by trans fats.

As seen below, the excess weight in the TRANS group was primarily in the form of increased visceral fat:


An abdominal CT scan.  The lighter areas represent fat tissue.  Both pictures depict roughly similar amounts of fat in the outer region (subcutaneous fat), whereas the TRANS group had significantly more fat tissue within the viscera.

For reference:


In all, TRANS had 27% more fat mass.  Fasting glucose and insulin levels were unchanged but postprandial insulin levels were markedly elevated in TRANS (see below), suggesting that dietary trans fats indeed caused insulin resistance.  I boldly use the term “caused” because this was a fairly well-controlled intervention study; the only thing different between the groups was the diet.


The TRANS group gained a significant amount of fat mass despite an absence of excess calories.  This was most likely caused by the trans fat-induced insulin resistance and subsequent postprandial hyperinsulinemia.

It would appear as though trans fatty acids defied the laws of energy balance.  The TRANS group gained fat mass despite an absence of excess calories.  Even the most practical explanation bodes poorly for trans fatty acids… it would appear as though trans fats were capable of inducing nutrient anti-partitioning independent of food intake.

I can see two ways to interpret these data.

  1. Trans fatty acids have an independent effect on energy balance.  That is, they specifically reduce energy expenditure, which would make the initial 70 kcal/kg*day excessive.  This would account for the excess fat mass, compared to controls, but not necessarily the increased body weight.  455 kilocalories (70kcal/kg*d) should be sufficient to support a specific amount of body weight; it is difficult to imagine a scenario whereby muscle mass declined enough to significantly reduce metabolic rate to the point where 455 kilocalories was so excessive that fat mass increased significantly more than the amount of muscle lost.  IOW, if this possibility were true, I would have expected, at most, a similar body weight but more fat and less muscle, not simply way more fat.

Of course, these processes would occur simultaneously and discreetly in vivo, but for simplicity’s sake I’ve broken it down.

6.6 kg monkey, x 70 kcal/kg*d = 462 kcal/d

Loses 0.022 kilograms of muscle, new body weight = 6.578 kg… since FFM is reduced, BMR should be reduced.  It was a 0.33% loss of body weight which was entirely from muscle (in this theoretical example), so perhaps BMR declines proportionately 460.46 kcal/d (? there are more accurate formulas in the literature, but this approximation is sufficient for our purpose)

all of those excess calories formerly burned in the lost muscle are now available for storage in fat.

462 – 460.46 = 1.54 excess kcal/day.  1.54 excess kcal every day for 6 years = 3,372.6  total excess kcal, which translates to ~0.438 kg fat mass.

0.438 kg new fat mass – 0.022 kg muscle lost = 0.416 kg overall weight gain.

6.6 kg + 0.416 = 7.016 kg.  Actual final body weight was 7 kg.  Pretty darn close.

Wow, can the loss of less than one ounce of muscle really cause such a drastic change in fatness?!? I don’t know for sure, but exchanging the microwave popcorn for a little resistance exercise seems prudent.

(in case you were wondering, no. I didn’t guess 22 grams. I did a ton calculations to quantify the metabolic rate reduction necessary to cause an energy surplus big enough to lay down enough fat mass to compensate for the reduction in muscle [which theoretically declined in proportion to the reduction in metabolic rate] and end up as close to 7 kg as possible… it could be calculated exactly but this has taken up 30 minutes already, and I think the point has been made)

2. Alternatively:  Energy expenditure varies day-to-day, hour-to-hour, second-to-second.  When we eat, we are transiently in positive energy balance, which reverses after a few hours, especially at night when a negative energy balance ensues and the fuel stored during the positive energy balance is utilized.  During those stints of positive energy balance, some of the excess energy is stored as fat tissue, while the rest is used to fuel the body.  Somehow, trans fatty acids shift the balance in favor of fat storage.

2a.  can there exist a positive energy balance selectively in adipose tissue?

2b. more likely, trans fatty acids reduce some component of energy expenditure, possibly basal metabolic rate, or perhaps the thermic effect of feeding.  Neither of these was measured, but I firmly believe energy balance was maintained.  It’s always maintained.

But the frightful conclusion remains the same: the TRANS group got fatter without eating more.  They didn’t eat more than they were supposed to but got fatter anyway.  Sad but true.

What about in humans?

Effect of trans-fatty acid intake on insulin sensitivity and intramuscular lipids-a randomized trial in overweight postmenopausal women. (Bendsen et al., 2011 Metabolism)

This study gave a group of 52 overweight but otherwise healthy postmenopausal women 16 grams of trans fatty acids in pumpkin muffins.  The control group received olive & palm oil-enriched pumpkin muffins.  In terms of the dosing, this study is almost directly comparable to Kavanaugh’s study.  However, this study only lasted 16 weeks (probably due to ethical reasons).  They also included a lean control group (for good measure?), baseline subject characteristics are below:


Nothing out of the ordinary.

And the investigators measured compliance empirically.  You are what you eat.  When a specific type of fat is consumed, its constituent fatty acids accumulate in body tissues like adipose and red blood cells.  So the researchers measured red blood cell trans fatty acid content.  Kudos!  (biomarkers are superior to almost any other measurement of compliance to a dietary intervention in humans)


Indeed, the women ate their muffins.  But no effect on body weight!


Body weight increased by about 2% in both groups.  If you want to get nit-picky, then we can make a few verrry long stretches concerning the body composition data:


The increase in fat mass was 33% greater in TFA compared to controls!  (fat mass increased by 3% in the control group and 4% in TFA).  The increase in percent body fat was twice as big in TFA compared to controls!  (body fat percent increased by 1% in the control group and 2% in TFA).  IOW, the changes in body composition were nil.  This does not necessarily refute Kavanaugh’s African green monkeys because that study lasted 6 years; the insignificant changes in fat mass in Bendsen’s women over the course of 16 weeks could very well add up to significant changes after 6 years.  Actually, if the endpoint was indeed a 6% weight gain after 6 years (like the monkeys) (78.7 * 1.06 = 83.422 -78.7 = 4.722 / 6 years = 2.156 grams per day x 16 weeks = 241 grams) we might have expected these women to gain less than they actually did (~ 241 grams in 16 weeks compared to 1,200 grams).  In truth, however, these numbers are well beneath what can actually be measured accurately even in a laboratory setting.  So it is bona fide nit-picking.

Maybe it’s time to throw in the towel and confess that trans fats are significantly worse for African green monkeys than for overweight but otherwise healthy postmenopausal women.  There was no change in visceral adipose between the groups:


Potential confounding?  I’m really grasping at straws… but here it goes anyway:


The TFA group reduced their carbohydrate consumption over the course of the study.  Carbohydrate consumption is directly correlated with liver fat accumulation.  But as per the trans fat study, trans fat consumption is inversely correlated with liver fat.  So we might expect trans fat-induced increase in liver fat to be cancelled out by the carb reduction-induced decrease in liver fat.  And this is exactly what happens!


Furthermore, liver fat didn’t change in the control group because 1) their carbohydrate intake didn’t decline, and 2) they weren’t eating a ton of trans fats.  IOW, neither of the major dietary determinants were altered.

So, according to these colorful explanations, trans fats may have been just as harmful to Bendsen’s women as they were to Kavanaugh’s monkeys.  The reason why the results differ can be at least partially explained by the inferior dietary intervention utilized by Bendsen.  IOW, Kavanaugh’s dietary intervention was perfect; the subjects (monkeys) ate their prescribed diet exactly, no cheating, no sneaking in any snacks.  The diet changed markedly in Bendsen’s study; all women gained weight meaning that the test foods were probably not isocalorically substituted for foods in their normal diet.  Perhaps they just ate the foods in addition to their normal foods (unlikely considering the marked changes in macronutrient consumption).

Some more data from Bendsen’s overweight but otherwise healthy postmenopausal women were reported in another paper, and hints of trans fat-induced insulin resistance were revealed…

Here are the results from an oral glucose tolerance test:


Glucose:  Just like Kavanaugh’s monkeys, there was no change in the glycemic response to a glucose load.

On the right, insulin levels.


Insulin:  The open triangles are the control group, solid squares are the TFA group..  The dashed lines indicate insulin responses at baseline and the solid lines represent insulin responses at 16 weeks. In a randomized controlled intervention study, ALL changes in the intervention group must be compared not only to baseline measurements (pre-treatment), but more importantly they must be compared to the changes in the control group.  Over the 16 weeks, insulin response declined very slightly in the control group (see the little red arrow around the 45 minute mark).  However, insulin response increased slightly in the TFA group.  Take either of these changes individually and they would amount to nil.  But when you consider the change in control vs that in TFA, a modest trend appears.  The TFA group is beginning to show a hint of peripheral insulin resistance.  Maybe I’m seeing something where there is nothing, but the Kavanaugh’s study lasted 6 years and Bendsen’s study was only 16 weeks.  We must expect the changes to be ~20 times smaller in the Bendsen study.

OK, perhaps I got lost in the minutiae, or lost sight of the forest for the trees, but that doesn’t mean I’ll be eating microwave popcorn any time soon.  And on the bright side, creating this post was a great brain exercise

calories proper


















attention nutrition researchers

how not to do a diet study.

As previously blogged about here, pair-feeding is an interesting phenomenon.

A high oxidised frying oil content diet is less adipogenic, but induces glucose intolerance in rodents. (Chao et al., 2007 British Journal of Nutrition)

Basically, these researchers wanted to test the effects on body weight and glucose tolerance of soybean oil that was used for deep frying, like for French fries.  WRT diet and food intake, the study was well designed.  There were four diets:

Divide and conquer

The “L” stands for “low,” as in Low SoyBean oil diet; this was the low fat control group.  The “H” stands for “high,” as in High SoyBean oil, High Oxidised oil, and High Fish oil.  Apparently, High Oxidised oil is not as delicious in rat chow as it is in French fries, so the rats fed HO ate considerably less.  But if the rats on HO ate less food, they would gain less weight and might exhibit improved glucose sensitivity compared to the other groups simply because of calorie restriction.  This would be a problematic confounding factor

Enter: pair-feeding.   In pair-feeding, the amount of HO ingested is regularly measured and an equivalent amount of calories are disbursed to the other groups, so that all groups are eating the same amount of calories.  Essentially, this controls for food intake so the effects of the diet can be tested directly, e.g., without being confounded by food intake.

As seen below, the pair-feeding regimen was successful:

The row outlined in red shows food intake.  Note: it is very similar in all four groups; this was due to pair-feeding.


For the purpose of clarity, and since I’m not concerned with what was actually being tested, here is a simplified table.



In red, the HO group ingested 369 kJ/d, so the HSB and HF groups were fed approximately 369 kJ/d.  However, look at the markedly different amounts of weight gained.  HO gained significantly less weight than HSB and HF despite eating the same amount.  Similarly, HSB gained less weight than HF despite similar food intake.  don’t jump down my throat just yet, there was no attempt to quantify energy expenditure in any group, but that doesn’t take away from my point.  Taken at face value, these data suggest a diet high in oxidized soybean oil hinders fat gain (regardless of the mechanism).

The researchers figured that if all the groups were fed ad libitum (could eat as much as they pleased), HO would gain less fat because they ate less (as opposed to a specific effect of the dietary fat composition, which was the question they wanted to address).  This was their justification for pair-feeding.

Since HO gained less fat despite pair-feeding, their first point was proven.  Therefore, the researchers discarded the difficult and labor-intensive pair-feeding for experiment #2.  As seen below, rats fed HO ad lib do indeed ingest less than HSB.  AND they gain less weight.


So my question is: did pair-fed HSB rats gain more fat than HO because of an anti-fattening effect of HO, or because of the stress imposed by being pair-fed???  Normally, ad libitum feeding occurs all throughout the night (lots of small meals).  When an animal is pair-fed, they are given the food all at once and they gorge because: 1) they are being under-fed (given less than for what they are hungry); and 2) they don’t know when will be their next meal.

So back to my question: for what exactly does pair-feeding control?  in the first experiment, calories were the same, but HO were ingesting oxidized oils while HSB had a stressful feeding regimen… there are two variables.  The results from experiment #2 show nothing but what we’d expect, i.e., eat less = gain less fat.  My point is simple: pair-feeding might control for one problem but it introduces another.  If there are any nutrition researchers reading, please consider this.

Their experiments therefore specifically do NOT address the question they asked.  Maybe it was good enough for the British Journal of Nutrition, but it tells us nothing definitive about “the adipogenic effect of high oxidized frying oil.”

For ways around this, and to learn how to design a much better experiment, I am happy to consult for a small fee :)     for some more free background information, read on:

The effect of feeding frequency on diurnal plasma free fatty acids and glucose levels. (Bortz et al., 1969 Metabolism)

In this experiment, they fed young healthy men the following diet divided into one meal per day, 3 meals per day, or 9 meals per day.  Rodents feed all night long, so they would be most similar to the men being fed 9 times per day.  A pair-fed rodent, on the other hand, is fed only once and they eat everything in one sitting, just like the men in this study who were given all their calories at once.

The differences in blood glucose and free fatty acid responses to the meal were robust:


Hyoooge differences in serum glucose and free fatty acids.

So in addition to the stress-inducing nature of being pair-fed, there are also profound physiological differences in nutrient handling which most likely contribute to differential fuel partitioning.

For more examples of how a restricted feeding regimen can go terribly wrong, see here and here.

When you consider the possibility that the act of pair-feeding can have distinct metabolic effects, independent from whatever intervention is being administered, the results become increasingly difficult to decipher.

Effects of pair-feeding and growth hormone treatment on obese transgenic rats (Furuhata et al., 2002 European Journal of Endocrinology)

In brief, there were 3 groups: control, transgenic growth hormone-expressing mice (who eat considerably more than control mice, abbreviated “TG”), and transgenic growth hormone-expressing mice pair-fed with control.  As seen below, the pair-fed group ate (by design) and weighed just as much as control:



But when looking at the metabolic profiles of these mice, things get somewhat complicated:


Lets start out with the second row, FFA. OK, so TG mice eat more, weigh more, and have higher FFA than control (1.34 vs. 0.88 mM).  The elevated FFA could be caused by: 1) increased food intake; 2) increased body weight; or 3) the transgene.  To determine if it was caused by option #3, we look to the TG/pair-fed group.  If FFA are similar to TG, then it was caused by the transgene (option #3).  If FFA are similar to control then it was caused by options #1 or #2.  Alas, FFA in the TG/pair-fed group are similar to TG suggesting it is a specific of the transgene, independent of food intake body weight.

However, if we take a look at the first row, triglycerides, it is not so clear.  Again, we know that TG mice eat more, weigh more, and have higher triglycerides than control.  The elevated triglycerides could be caused by: 1) increased food intake; 2) increased body weight; or 3) the transgene.  To determine if it was caused by option #3, again we look to the TG/pair-fed group.  If triglycerides are similar to TG then it was caused by the transgene (option #3).  If triglycerides are similar to control, then it was caused by options #1 or #2.  But oh no!  Triglycerides in the TG/pair-fed group are significantly lower than TG and control!  This means we have to consider the fourth option: it was caused by pair-feeding.

The other two rows are easier to interpret, as insulin and leptin in the TG/pair-fed group exhibited an intermediate phenotype between control and TG, suggesting they were partially mediated by downstream effects of the transgene (i.e., increased food intake or body weight).  In the conclusion of the paper, the authors aptly explain the changes in insulin and leptin but cleverly avoid the triglycerides.  In their favor, the depressed triglycerides could have been an artefact, but the possibility that they were a product of the pair-feeding per se cannot be ruled out.  So depending on the question being addressed, pair-feeding has the potential to royally screw things up.


Calories proper