Overfeeding Studies

There are quite a few interesting overfeeding studies. The typical format of these studies is to take subjects and first determine the caloric intake to keep them in energy balance (weight stable). The study will then increase one of the three macronutrients (fat, carbohydrate or protein) and then look at the effects. Often studied are fat accumulation, body composition changes or blood lipids. The change is then attributed to the changed macronutrient.

Both fat and carbs are shown to increase weight and make body composition worse. Protein has been shown to not increase weight and increased protein improves body composition.

The problem I have with the study methodology is that the changes can’t all be attributed to the increased macronutrient alone since the increased macronutrients interacts with the other baseline macronutrients. 

Take as an example a baseline diet which provides 25% of calories from protein, 50% from carbohydrates and 25% from fat. For a 2000 calorie a day person that’s 500 calories from protein (125g), 1000 calories from carbohydrates (250g) and 500 calories from fat (55.6g).

Adding 500 calories a day of fat, for instance would change the fat from 500 calories to 1000 calories (111g) while leaving the carbohydrates and protein at the same amounts. So if there’s an increase in body fat how much of a change in body fat can be attributed to the fat alone? How much of the increase in weight is due to the interaction between fat and, say, carbohydrates? Perhaps there’s a carbohydrate/fat limit where if you exceed the amount of carbs/fat it causes much more fat storage due to the combination of the two?

Low Carb Diet Reduces the Variables

The Low Carb Diet essentially reduces the three [macronutrient] variables to two. The calories from carbohydrates are typically 5% on a very low carb diet. So the only two macronutrients left are protein and fat. Yet, even this is no guarantee for weight loss. There are people (think Jimmy Moore) who eat on the very low protein and high fat end. There are others (think Ted Naiman) who eat on the higher protein and lower fat end.

Jimmy and Ted

Clearly, what Jimmy is doing isn’t working well for Jimmy and what Ted is doing is working quite well for Ted. Jimmy is an n=1 for overeating fat. Ted is an n=1 for eating more protein. However, Ted controls for total calories and Jimmy seems to have no clue how many calories he eats in a particular day.

The difference may be exercise. It may be diet. I have a hard time finding a picture of a higher protein advocate who looks like Jimmy but I can find plenty of keto personalities who eat a lot of fat and look more like Jimmy (the Two Keto Dudes comes to mind).

A Fear of Protein?

Jimmy and others have been afraid of protein with the fear that eating protein causes the protein to turn to chocolate cake (Jimmy is infamous for making the comparison to chocolate cake at one point). I’ve looked at this subject in many posts in this BLOG (Protein does not turn into chocolate cake). 

Ted says he spends most of his day trying to convince diabetics that they should eat more protein. It is true that protein does raise blood sugar by a small amount in a diabetic but the benefits outweigh that small rise and if a person is not a diabetic protein will lower blood sugar (Glucose Response to Protein).

Problem with Protein Studies

Protein studies are used to determine protein requirements. These studies look at nitrogen balance which is either negative (the person isn’t getting enough nitrogen from their diet) or positive (the person is getting enough nitrogen from their diet). 

The problem is that protein studies are based on so-called “balanced” diets where carbohydrates are available to make the amount of glucose required by the body (Low Carbs and Gluconeogenesis). These studies don’t include the effects of gluconeogenesis (GNG). If you are eating low carb then protein provides the substrate materials (from your diet) for (GNG). For diabetics their body is already really good at making glucose via GNG (Gluconeogenesis – Later Thoughts). 

If you barely eat enough protein to meet the minimum (nitrogen replacement) requirements then your body will get it’s GNG needs from fat. Suppose that the body requires 120g of carbohydrates per day for the brain and other essential organs. If you eat 20g of carbohydrates a day that’s 100 short. If half the protein gets converted to glucose and your body requires 200g of protein to provide that glucose. (Note these are very rough numbers but the idea applies).

So, if you are on a low carbohydrate diet you need more protein than just your replacement needs. You also need protein to meet your GNG needs.

Many Ways to Lose Weight

There are quite a few ways to lose weight. Most of them involve eating less calories than you burn. You can lose weight with Low Fat or Low Carb diets. You can even lose weight with a Low Protein diet. All of these work if you are at a caloric deficit. High fat and high carbs at the same time don’t work at the same time unless your goal is weight gain.

Also, there’s an interaction with the macronutrient type. Some macronutrients encourage fat gain. Truthfully, fat is always stored easily as fat – but only accumulates in a caloric surplus. You burn off what you eat if you are in fat balance. If you eat less fat you lose body fat. If you eat more fat you will gain body fat.

Even people who eat a carnivore diet are eating a large portion of their calories as fat. As an example: Ribeye Steak from Walmart has 22g of protein and 20g of fat in an 4 oz serving. That’s 88 calories from protein and 180 calories from fat. Or 33% of calories from protein and 67% of calories from fat.

Consuming large amounts of fat is unavoidable in the weight maintenance portion of Low Carb diets. There’s a top limit on the amount of protein that you can/should eat. As an example, if you are eating 1 g of protein per lb of body weight and you weigh 200 lbs that’s 200 gram of protein or 800 calories. The rest of your daily caloric needs will then come from fat.  If you are eating 2000 calories a day that’s 1200 calories from fat.

The problem is that many people, like Jimmy Moore, eat at a maintenance or higher level of calories and macros when they need to be in a weight loss phase. Jimmy eats low protein and carbs so it’s not protein or carbs that are making Jimmy fat. The extra fat that Jimmy eats accumulates as fat when he eats more calories from fat regardless of whether he avoids carbohydrates or not. Fat doesn’t magically vanish when you eat it and it doesn’t require much energy to store -it is about 95% efficient to store fat (Eat Too Much Fat – Get Fat).

Maximizing Weight Loss

The only way for someone like Jimmy to lose weight is to get enough protein and limit dietary fat. If Jimmy has a goal weight of 200 lbs he should eat 200 grams of protein spread over four meals a day of 50 grams per meal  (Protein Gurus – Part 2). That maximizes Muscle Protein Synthesis and provides enough substrate to maintain his blood glucose. Jimmy should then eat enough fat to cover the amount he won’t be eating from his body. Jimmy has at least 100 lbs of fat mass and could easily have a 3000 calories a day deficit. It wouldn’t be at all pleasant but he could do it.

If you want to find out what you can do on a maximum fat loss diet, check out our Keto calculator.

Improved Overfeeding Studies

So how would you improve an overfeeding study? I would isolate the macronutrients and absolutely minimize the other macronutrients. Do a lean protein study (essentially a PSMF study) with variable protein levels. I’d do a carbohydrate variation study with minimal protein and fat. I’d do a fat level study with minimal carbs and protein. All of them in isolation. Wouldn’t be a very balanced diet at 90%, 5%, 5%. Couldn’t be too long a term. May not pass ethics boards. But it could tease out the interactions between the macronutrients. 

Fructose – Part 2

Here’s another great paper on the role of Fructose (Lisa C Hudgins; Why do sweets fatten our livers?, The American Journal of Clinical Nutrition, Volume 96, Issue 4, 1 October 2012, Pages 685–686).

Here is another study (Sevastianova K. Effect of short-term carbohydrate overfeeding and long-term weight loss on liver fat in overweight humans
Am J Clin Nutr. 2012 Oct;96(4):727-34. Epub 2012 Sep 5.).

Carbohydrate overfeeding for 3 wk induced a >10-fold greater relative change in liver fat (27%) than in body weight (2%). The increase in liver fat was proportional to that in DNL. Weight loss restores liver fat to normal.

These data indicate that the human fatty liver avidly accumulates fat during carbohydrate overfeeding and support a role for DNL in the pathogenesis of NAFLD.

Testing the Protein Leverage Hypothesis

There’s the common view of obesity that it’s due to increased fat and/or carbs in the American diet. And the statistics bear out that increase (Gregory L Austin, Lorraine G Ogden, James O Hill; Trends in carbohydrate, fat, and protein intakes and association with energy intake in normal-weight, overweight, and obese individuals: 1971–2006, The American Journal of Clinical Nutrition, Volume 93, Issue 4, 1 April 2011, Pages 836–843):

The prevalence of obesity increased from 11.9% to 33.4% in men and from 16.6% to 36.5% in women. The percentage of energy from carbohydrates increased from 44.0% to 48.7%, the percentage of energy from fat decreased from 36.6% to 33.7%, and the percentage of energy from protein decreased from 16.5% to 15.7%.

There’s an interesting note:

In NHANES 2005–2006, a 1% increase in the percentage of energy from protein was associated with a decrease in energy intake of 32 kcal (substituted for carbohydrates) or 51 kcal (substituted for fat).

What is the Protein Leverage Hypothesis?

The central claim is that protein is being displaced by increasing amount of carbs and fat. From this paper (Alison K. Gosby , Arthur D. Conigrave, Namson S. Lau, Miguel A. Iglesias, Rosemary M. Hall, Susan A. Jebb, Jennie Brand-Miller, Ian D. Caterson, David Raubenheimer, Stephen J. Simpson. Testing Protein Leverage in Lean Humans: A Randomised Controlled Experimental Study. PLoS ONE 6(10): e25929.):

The ‘protein leverage hypothesis’ proposes that a dominant appetite for protein in conjunction with a decline in the ratio of protein to fat and carbohydrate in the diet drives excess energy intake and could therefore promote the development of obesity.

The study found:

In our study population a change in the nutritional environment that dilutes dietary protein with carbohydrate and fat promotes overconsumption, enhancing the risk for potential weight gain.

Here’s the chart showing the differences:

From the study:

Simpson and Raubenheimer (Simpson, S. J. and Raubenheimer, D. (2005), Obesity: the protein leverage hypothesis. Obesity Reviews, 6: 133-142.) used data from the FAOSTAT [5] nutrient-supply database to show that an estimated decrease in percent dietary protein from 14% to 12.5% between 1961 and 2000 in the USA was associated with a 14% increase in non-protein energy intake, with absolute protein intake remaining almost constant.

Not All [Sugar] Is Bad

It turn out that not all sugar is bad. Put another way, not everything in sugar is bad for diabetics. Sugar (sucrose) consists of one glucose and one fructose molecule, or 50% glucose and 50% fructose. The body does different things with glucose vs fructose.

There are several studies which tease out the differences between glucose and fructose. Here’s one of the studies (Kimber L. Stanhope, et.al. Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/obese humans. The Journal of Clinical Investigation, 2009;5:119, pp 1322-1334). The study:

To assess the relative effects of these dietary sugars during sustained consumption in humans, overweight and obese subjects consumed glucose- or fructose-sweetened beverages providing 25% of energy requirements for 10 weeks.

Switching out between the two sugars sounds like a fairly easy test and should have resulted in identical results. And the results were the same for weight gain so both parts of sugar can make you fatter.

There was a very important difference, though, where the fat was located. In the group eating fructose the visceral adipose volume was significantly increased only in subjects consuming fructose. 

Fasting plasma triglyceride concentrations increased by approximately 10% during 10 weeks of glucose consumption but not after fructose consumption.

In contrast, hepatic de novo lipogenesis (DNL) and the 23-hour postprandial triglyceride AUC were increased specifically during fructose consumption. Similarly, markers of altered lipid metabolism and lipoprotein remodeling, including fasting apoB, LDL, small dense LDL, oxidized LDL, and postprandial concentrations of remnant-like particle–triglyceride and –cholesterol significantly increased during fructose but not glucose consumption.

In addition, fasting plasma glucose and insulin levels increased and insulin sensitivity decreased in subjects consuming fructose but not in those consuming glucose.

These data suggest that dietary fructose specifically increases DNL, promotes dyslipidemia, decreases insulin sensitivity, and increases visceral adiposity in overweight/obese adults.

Interesting results. For a newer paper which summarized other studies on the subject see this (Stanhope KL, Schwarz J-M, Havel PJ. Adverse metabolic effects of dietary fructose: Results from recent epidemiological, clinical, and mechanistic studies. Current Opinion in Lipidology. 2013;24(3):198-206.)

Saturated Fat and Liver Fat

There’s a study which indicates that PUFAs protect against accumulation of liver fat but SFA (saturated fatty acids) contribute to the accumulation of liver fat (Fredrik Rosqvist, David Iggman, Joel Kullberg, Jonathan Cedernaes, Hans-Erik Johansson, Anders Larsson, Lars Johansson, Håkan Ahlström, Peter Arner, Ingrid Dahlman, Ulf Risérus. Overfeeding Polyunsaturated and Saturated Fat Causes Distinct Effects on Liver and Visceral Fat Accumulation in Humans. Diabetes Jul 2014, 63 (7) 2356-2368). The study overfed young subject muffins for seven weeks made with PUFA or SFA. The SFA participants gained fat in their liver and the PUFA group did not gain liver fat.

However, there’s one interesting point in the study:

In the current study, a fructose–SFA interaction on liver fat is possible since the muffins contained significant amounts of fructose. Early animal data showed that carbohydrate-induced lipogenesis was inhibited by adding linoleic acid, whereas palmitate had no effect, and SFAs have enhanced steatosis and increased hepatic lipogenesis compared with PUFAs.

That makes a lot of sense. 

The PUFA was Sunflower Oil. The SFA was Palm Oil. It would have been interesting if the SFA was plant based.

Alcohol and Blood Sugar

It is widely known that alcohol lowers blood sugar, but why does it? From an interesting study (CHARLES U. LOWE, LUIS L. MOSOVICH. The Paradoxical Effect of Alcohol on Carbohydrate Metabolism in Four Patients with Liver Glycogen Disease. Pediatrics, June 1965, VOLUME 35 / ISSUE 6.).

Oxidation of ethanol to acetaldehyde, the first step in alcohol metabolism, is catalyzed by the enzyme alcohol dehydrogenase and results in the reduction of DPN to DPNH. In a coupled reaction, pyruvate is converted to lactate with regeneration to DPN. There are a number of consequences of these reactions when alcohol is consumed. Lactate levels in blood rise; DPNH produced by the reaction inhibits the enzymatic steps involved in the conversion of UDP galactose to UDP glucose and glutamate to alpha keto glutarate. As a result of these inhibitions, galactose removal from blood is markedly delayed and gluconeogenesis from amino acids is inhibited.

This could help explain why people report easily getting plastered on Low Carb diets. A lack of dietary carbohydrates means that someone on a Low Carb is producing their blood glucose through Gluconeogenesis (GNG). If alcohol inhibits GNG then blood sugar may drop farther on a Low Carb diet than on a Higher Carb diet.

Protein Contradictions

Much of the popular press writes that we should eat more meals a day. As an example (How Much Protein for Strength and Mass Gains?):

total protein amount should be spread out over 5 to 6 intakes a day

They advise the amount of protein to be:

For males, who aim at increasing muscle mass and strength gains, if you only train once a day, 2 g a kg should be more than enough (for women 1.2g /kg of bodyweight).

Let’s do the math here. Suppose someone is 75 kg (about 165 lbs). At 2g/kg that would be 150 grams of protein per day. If they eat 5 meals a day that would be 30 grams of protein per meal. The problem is that they will probably not ever reach the Leucine threshold at any of the meals (Protein Gurus – Part 2). As a result they will never maximize muscle protein synthesis.

Also the timing between protein meals should be 5 hours and that would be 25 hours of eating in a day. Doesn’t quite fit.

My current optimized method is three protein meals a day spread out by five hours (Muscle Protein Synthesis Meal Spacing Maximum). This can be challenging and does require advance planning for meals.


Revisiting the “Alpert” Number

I think that the Alpert number may not be right if you are on a Low Carbohydrate diet.

The Alpert number is the maximum rate of fat oxidation from a relatively moderately active person (Hypophagia – How much fat can I lose in a day?). It occurred to me that I can check this number from my own VO2max test.

  • Looking at the REE at rest (REE from VO2max) it shows 2.16 kCal/min.
  • From my Bod Pod results (Overshot My Recomp Goals – Part 1) my fat mass is 12.3 lbs.
  • Multiplying my fat mass times the Alpert number is 381.3 kCals/day. That’s 15.88 kCal/hr or 0.26 kCal/min.

Yet, my REE was 2.1 kCal/min  at and RER of 0.73 (90% fat) which is 1.9 kCal/min from fat oxidation. Flipping the number around that’s 1.9 times 60 times 24 = 2736 kCal per day from fat.

The smallest number I saw in the resting period was 1.209 kCal/min or 1740 kCal/day. dividing 1740 number by my fat weight in lbs is 141 kCal per lb of fat mass. That’s quite a bit more than the Alpert number.

The Minnesota Starvation (Ancel Keys) data was the basis of the Alpert number. Perhaps the difference is in the idea that I am not actually in starvation? And the Minnesota Starvation subjects were fed carbohydrates in their diet.

The Alpert number pretty closely matches my own experiences in Protein Sparing dieting.

Is the Magic in Low Carb, High Fat?

How about neither? How about if the magic is in the increased protein content of LCHF diets? That was the question that this study sought to unravel (Stijn Soenen, Alberto G. Bonomi, Sofie G. T. Lemmens, Jolande Scholte, Myriam A. M. A. Thijssen, Frank van Berkum, Margriet S. Westerterp-Plantengaab. Relatively high-protein or ‘low-carb’ energy-restricted diets for body weight loss and body weight maintenance? Physiology & Behavior. Volume 107, Issue 3, 10 October 2012, Pages 374-380.).


Body-weight (BW), fat mass (FM), blood- and urine-parameters of 132 participants (age = 50 ± 12 yr; BW = 107 ± 20 kg; BMI = 37 ± 6 kg/m2; FM = 47.5 ± 11.9 kg) were compared after 3 and 12 months between four energy-restricted diets with 33% of energy requirement for the first 3 months, and 67% for the last 9 months: normal-protein normal-carbohydrate (NPNC), normal-protein low-carbohydrate (NPLC); high-protein normal-carbohydrate (HPNC), high-protein low-carbohydrate (HPLC); 24 h N-analyses confirmed daily protein intakes for the normal-protein diets of 0.7 ± 0.1 and for the high-protein diets of 1.1 ± 0.2 g/kg BW (p < 0.01).


BW and FM decreased over 3 months (p < 0.001): HP (− 14.1 ± 4 kg; − 11.9 ± 1.7 kg) vs. NP (− 11.5 ± 4 kg; − 9.3 ± 0.7 kg) (p < 0.001); LC (− 13.5 ± 4 kg; − 11.0 ± 1.2 kg) vs. NC (− 12.3 ± 3 kg; − 10.3 ± 1.1 kg) (ns). Diet × time interaction showed HPLC (− 14.7 ± 5 kg; − 11.9 ± 1.6 kg) vs. HPNC (− 13.8 ± 3 kg; − 11.9 ± 1.8 kg) (ns); NPLC (− 12.2 ± 4 kg; − 10.0 ± 0.8 kg) vs. NPNC (− 10.7 ± 4 kg; − 8.6 ± 0.7 kg) (ns); HPLC vs. NPLC (p < 0.001); HPNC vs. NPNC (p < 0.001). Decreases over 12 months (p < 0.001) showed HP (− 12.8 ± 4 kg; − 9.1 ± 0.8 kg) vs. NP (− 8.9 ± 3 kg; − 7.7 ± 0.6 kg) (p < 0.001); LC (− 10.6 ± 4 kg; − 8.3 ± 0.7 kg) vs. NC (11.1 ± 3 kg; 9.3 ± 0.7 kg) (ns). Diet × time interaction showed HPLC (− 11.6 ± 5 kg ; − 8.2 ± 0.7 kg) vs. HPNC (− 14.1 ± 4 kg; − 10.0 ± 0.9 kg) (ns); NPNC (− 8.2 ± 3 kg; − 6.7 ± 0.6 kg) vs. NPLC (− 9.7 ± 3 kg; − 8.5 ± 0.7 kg) (ns); HPLC vs. NPLC (p < 0.01); HPNC vs. NPNC (p < 0.01). HPNC vs. all other diets reduced diastolic blood pressure more. Relationships between changes in BW, FM, FFM or metabolic parameters and energy percentage of fat in the diet were not statistically significant. Metabolic profile and fat-free-mass were improved following weight-loss.

Also (A. K. Gosby A. D. Conigrave D. Raubenheimer S. J. Simpson. Protein leverage and energy intake. Etiology and Pathophysiology, 28 October 2013).

…these trials encompassed considerable variation in percent protein (spanning 8–54% of total energy), carbohydrate (1.6–72%) and fat (11–66%). The data provide an opportunity to describe the individual and interactive effects of dietary protein, carbohydrate and fat on the control of total energy intake. Percent dietary protein was negatively associated with total energy intake (F = 6.9, P < 0.0001) irrespective of whether carbohydrate (F = 0, P = 0.7) or fat (F = 0, P = 0.5) were the diluents of protein. The analysis strongly supports a role for protein leverage in lean, overweight and obese humans.

Keto for the Win – Again

Here is another cross-over study showing the advantage of the keto diet over a medium carbohydrate diet (Johnstone AM, Horgan GW, Murison SD, Bremner DM, Lobley GE. Effects of a high-protein ketogenic diet on hunger, appetite, and weight loss in obese men feeding ad libitum. Am J Clin Nutr. 2008 Jan;87(1):44-55.).

Ad libitum energy intakes were lower with the LC diet than with the MC diet [P=0.02; SE of the difference (SED): 0.27] at 7.25 and 7.95 MJ/d, respectively. Over the 4-wk period, hunger was significantly lower (P=0.014; SED: 1.76) and weight loss was significantly greater (P=0.006; SED: 0.62) with the LC diet (6.34 kg) than with the MC diet (4.35 kg). The LC diet induced ketosis with mean 3-hydroxybutyrate concentrations of 1.52 mmol/L in plasma (P=0.036 from baseline; SED: 0.62) and 2.99 mmol/L in urine (P<0.001 from baseline; SED: 0.36).

These men were allowed to eat as much as they wanted but chose to eat less when they were given Low Carb food.