More Proof: FAT FATTENS BEST

Fat vs. Carbohydrate Overeating: Which Causes More Fat Gain?

Two human studies, published in 1995 and 2000, tested the effect of 
carbohydrate vs. fat overfeeding on body fat gain in humans.  

What did they find, and why is it important?

We know that daily calorie intake has increased the US, in parallel with the dramatic 
increase in body fatness.  These excess calories appear to have come from fat, 
carbohydrate, and protein all at the same time (although carbohydrate increased the most).  
Since the increase in calories, carbohydrate, fat, and protein all happened at the same time, 
how do we know that the obesity epidemic was due to increased calorie intake and not just increased carbohydrate or fat intake?  If our calorie intake had increased solely by the addition of 
carbohydrate or fat, would we be in the midst of an obesity epidemic?

The best way to answer this question is to examine the controlled studies that have compared carbohydrate and fat overfeeding in humans.

Horton et al.

The first study to address this question was published in 1995 by Dr. James Hill's group (1).  
The title of the paper is "Fat and Carbohydrate Overfeeding in Humans: Different Effects 
on Energy Storage".  Sounds promising for the calorie skeptics.  

Sixteen men (9 lean*, 7 obese) were overfed by 50 percent of calorie needs for 14 days, 
by adding excess fat or carbohydrate to the diet.  After a four-week break, each person 
was overfed the macronutrient they had not received the first time (randomized crossover design).  Subjects were provided with all food,  in a research kitchen, although they were allowed
 to consume some of it at home.

After 14 days, the researchers measured changes in body weight, fat mass, and lean mass. 

Here's what they found:

There were no significant differences between diets and/or groups in body weight or body composition changes.  

Carbohydrate and fat overfeeding caused nearly identical increases in body weight, fat mass,
 and lean mass, and this was true both in the lean and obese groups.  

Here's a graph of body weight changes:

And here's a graph of changes in fat mass:

Note that obese subjects seemed to gain more fat than lean subjects.  This is presumably because they were overfed by a greater absolute number of calories**.

If changes in body fatness were essentially identical during fat and carbohydrate overfeeding, why did the investigators choose to state in the title that "energy storage" differed between macronutrients?  The reason is that carbohydrate overfeeding caused an increase in carbohydrate burning and total energy expenditure, while fat overfeeding had no significant effect on fat burning or total energy expenditure.  In other words, the body "burned off" some of the extra carbohydrate, but it didn't burn off any of the extra fat.  A higher proportion of the fat calories was stored as body fat.

Yet in the end, the differences were small-- the body stored nearly all of the excess calories in both cases, and any apparent differences in energy expenditure were not reflected in fat mass***.  Calorie-for-calorie, body fat accumulation was approximately the same during fat and carbohydrate overfeeding.

Lammert et al.

The second study was published in 2000 by the research group of Dr. Bjorn Quistorff (2).  Ten pairs of lean young men were overfed by 5 megajoules (1,195 kcal) per day for 21 days, given as either a carbohydrate-rich or a fat-rich diet****.  Subjects lived and ate in a research setting the entire time.  The study was extremely well controlled.

Body composition was determined weekly by underwater weighing.  Increases in body weight were similar between groups, and increases in fat mass were almost identical:

Interestingly, the carbohydrate-overfed group actually gained more lean mass than the fat-overfed group.  It's unclear to me whether that reflects actual tissue gain, or simply increased glycogen storage.  Another interesting thing to note is that fat gain varied tremendously between individuals.  Due to genetics, physical fitness, and/or other factors, some people simply store more body fat when they eat excess calories, while others burn most of it off.  This has been confirmed repeatedly.

In any case, this study confirms that fat gain is approximately the same whether people overeat fat or carbohydrate.

Bonus Study: Hirsch et al. 

We know that carbohydrate and fat cause approximately equal fat gain per unit calorie during overfeeding, but what happens when people aren't overeating?  Does the proportion of calories supplied as fat or carbohydrate affect body fatness in that scenario?

The research group of Drs. Rudy Leibel and Jules Hirsch kept a series of subjects under metabolic ward conditions, strictly controlling the diet and dramatically varying the proportion of carbohydrate to fat, while keeping calories constant, for several months at a time (3, 4).  Here's a summary of their findings:

We showed that the carbohydrate-to fat ratio could vary widely with little or no alteration in the energy requirement for weight maintenance.  The results of a 13-week study in which an individual was fed a formula diet extremely rich in carbohydrate and low in fat for a period of 38 d and, thereafter, for a longer time, a diet rich in fat and low in carbohydrate are shown in Figure 1 [see below-- SG].  Weight varied little throughout the study and average energy intake was the same throughout...  The reason for emphasizing these findings is that under the strict conditions imposed by hospitalization and feedings of a formula diet, energy needs are the same over long periods of time even though carbohydrate-to-fat ratios vary.  Similar data were accumulated in 15 subjects.

Here, you can see the weight trajectory of the woman described above:

Over a fairly long period of time, her weight stayed within a 1 kg range, despite huge differences in diet composition.  The same thing was observed in a number of other subjects.

In other words, under non-overfeeding conditions, the carbohydrate and fat content of the diet have no measurable impact on body weight when calories are controlled.

Conclusion

There are always caveats when interpreting scientific evidence.  Here are a few for today:

• The studies we discussed were small.
• They may not have been long enough for differences to emerge.
• The overfeeding studies didn't include women.
• The overfeeding studies reported significant individual variability.  In other words, some individuals may gain body fat more readily when they overeat carbohydrate, while others may be more sensitive to fat.  Or not.  We can't really say based on these studies, but it remains possible. 

Scientific evidence is never perfect, but at some point we have to decide which hypothesis is best supported.  In this case, the clear winner is the hypothesis that total calorie intake determines body fatness, but not the proportion of dietary fat or carbohydrate.  

Based on the available evidence, the US obesity epidemic probably resulted from an increase in total calorie intake, not from changes in fat or carbohydrate intake that were acting independently of total calories.  We would likely be faced with the exact same obesity epidemic if our increased calorie intake came exclusively from fat, or exclusively from carbohydrate.

That being said, macronutrients (fat, carb, protein) are not irrelevant to body fatness!  They can impact fat storage by affecting how many calories we eat.  Added fats tend to increase calorie intake, while high-protein diets tend to decrease calorie intake.  Of course, refined and processed versions of fat and carbohydrate tend to favor higher calorie intake as well, due in part to higher calorie density and palatability.

Somewhat paradoxically, once a person is overweight or obese, increasing the proportion of protein and fat at the expense of carbohydrate can help control appetite and reduce body fatness.  During moderate carbohydrate restriction, this effect seems to depend mostly on increased protein intake, but during more extreme carbohydrate restriction, there may be a role for ketones.  I think there are still mysteries here that deserve further scientific exploration.  

* Average body fat percentage of lean group = 21.4%, which is not particularly lean for a man.  BMI = 21.3, which is low, so these people probably had a fairly undesirable body composition on average.  The investigators specifically excluded "highly trained individuals".

** Subjects received a 50% calorie excess.  Since obese people have a higher baseline calorie expenditure than lean people (25% higher in this study), the excess calories during overfeeding would be greater in an absolute sense.

*** The investigators do state that the body composition changes they measured were near the detection limit, given the short duration of overfeeding.  The subtext is that the fat overfeeding group may have ended up gaining a bit more fat if the study had lasted longer.  Personally, I'm skeptical of this possibility, due to the nearly identical body composition changes they reported and the results of Lammert et al.

*** Carb-rich: 11%, 78%, 11% of calories as fat, carb, protein.  Fat-rich: 58%, 31%, 11% of calories as fat, carb, protein.  In the carb group, a lot of the calories came from sugar (sucrose).