Is a Calorie a Calorie? New study says, “No.”


A new study published in JAMA refutes the commonly held assumption that all calories are equal in terms of how they effect our metabolism (at least in a small group, n=21, of semi-starved obese subjects each rotating through different weight-maintenance diets for a month).  When subjects consumed an Atkin’s-like diet (Table 1: 10:60:30 % carb/fat/protein), they burned, on average, far more calories a day then when they were on a low-fat (60:20:20 % carb/fat/protein) and low glycemic index  diet (40:40:20 % carb/fat/protein) (Table 3: ~325 and ~200 calories a day more, on average, respectively) while on the SAME AMOUNT OF CALORIES with subjects exercising LESS on low carb diet than other two diets (also, the higher thermic effect of burning protein, 10% higher in Atkin’s, should have been offset by the higher carb% in the other two diets so this cannot explain the different metabolic rates).

This means when subjects were on the low-fat diet, they would have to exercise moderately for around an hour a day to meet their metabolic rate while on the Atkin’s diet: lead author, Dr. David Ludwig, explains in an MSNBC interview, “”That’s an amount equal to one hour of moderate physical activity, without lifting a finger.”  Researchers did not measure the weight or % body fat change by diets, possibly because four weeks was too short a time to see a change.  However, the increased metabolic rate on Atkin’s might lead to lower weight over-time compared with the less metabolically-active low-fat and low-glycemic index diets.  Lustig explains this metabolic difference according to the “insulin hypothesis of obesity” that the higher carbohydrate intake of the low-fat/GI diets raised insulin relative to Atkin’s which essentially lowered subjects’ metabolic rate, a common effect of starvation, as lead author, Ludwig remarked).  Lustig explains insulin, starvation mode, and why diets fail here.

Also, cholesterol measurements were better on Atkin’s then the other groups (i.e. HDL was higher and triglycerides were lower-meaning the group that ate the most fat had the least fat in their blood, a paradoxical but commonly observed effect of a very low carb, high fat diet relative to low-fat, high carb diet).  Subjects were also less insulin resistant on Atkin’s then the other diets (strangely, insulin resistance might be the primary underlying cause of metabolic syndrome, a precursor condition that might lead to many of the chronic diseases afflicting Westernized-countries like  hypertension>stroke, dyslipidemia>heart disease, hyperglycemia/hyperinsulemia>diabetes, cancer, Alzheimer’s, etc.).  Higher insulin sensitivity results in Atkin’s were measured as higher peripheral and hepatic insulin sensitivity levels which indicate, obviously, higher insulin sensitivity and less insulin resistance.  Leptin levels were also lower with Atkin’s, meaning risk of “leptin resistance,” a possible key factor in obesity and insulin resistance, might be lower with lower carb intake.  The only possible downside to Atkin’s was a slight elevation in stress (measured by cortisol) and inflammation (measured by c-reactive protein, or CRP) but these levels were both significantly lower than before the weight loss and well within healthy, “normal” ranges, at least according to (CRP, cortisol).  However, it is possible that subjects had not fully adapted to ketosis (burning fat and protein metabolites for fuel as opposed to glucose) and that this elevation was simply a temporary marker of this metabolic transition (this is simply a guess on my part, I could be wrong).

The lead author, Dr. David Ludwig, summarizes the results:

“The low-fat diet produced changes in energy expenditure and serum leptin42 – 44 that would predict weight regain. In addition, this conventionally recommended diet had unfavorable effects on most of the metabolic syndrome components studied herein. In contrast, the very low-carbohydrate diet had the most beneficial effects on energy expenditure and several metabolic syndrome components, but this restrictive regimen may increase cortisol excretion and CRP.”

My take:

These 21 obese subjects might be fundamentally different then other obese adults (now around 36% of the U.S. adult population), meaning that these results are not “representative” of most obese U.S. adults and therefore unimportant.  This may or may not be ture.

Regardless of whether these subjects represent most obese American adults who have dieted or not, in my opinion, this study is still hugely important because it likely provides a small glimpse into why dieting, in the conventional low-fat, high carb sense, is so hard, why so many adults yo-yo diet, and why most obese people regain lost weight (or even surpass their pre-diet weight).  This study might, if replicated in larger, longer future studies, explain why we have an obesity epidemic.

The simple explanation for what this study found is that eating calorically-identical diets of different nutrient composition (measured by carbs, proteins, and fats) led to different metabolic rates and different metbolic disease risks for the subjects.  Therefore, equal caloric intake of different nutrients might lead to different risks not only for cholesterol, inflammation, and insulin resistance, as most people accept as obvious, but for weight gain and, therefore, obesity as well.  It is this latter finding that is so radical and potentially useful, if these results are replicated in larger, longer studies (which is what is required in science for us to actually explain causality).

Therefore, the results suggest that if you are obese and want to lose weight and, more importantly, maintain weight loss, reduce the amount of sugar and refined carbohydrates you eat (possibly including whole grains) and replace them with fat and natural carbs (unlimited green veggies and moderate-GI fruit like berries and the occasional apple).  Eating processed foods like refined carbs (grains, starches, and sugar) elevate your blood sugar and insulin and damage your liver (fructose especially), which reduces your metabolic rate and increases your risk of metabolic syndrome.  Stop eating them, eat whole foods, and stop fearing fat, esp. saturated fats (because you have to replace the carbs with something and humans likely evolved to eat saturated fats so, if an animal’s ideal diet is eating what an animal evolved to eat, how could they be harmful?).

Lastly, everyone is different so even though the averages were better on low-carb for obesity and metabolic syndrome risk, a small number of people did better on a low-GI diet than low carb (see Fig. 3).  This means that while most obese people may be obese due to “carbohydrate intolerance” and would do better on an Atkin’s-like (or Paleo-style) diet than low-fat or low-GI, low-GI might be better for some.


Researchers semi-starved (“calorie-restricted”) 21 obese adults until they lost ~10-15% of their body weight.  Subjects were then assigned to eat three isocaloric (same # of calories) diets each with a different % composition of carbohydrates, fats, and proteins in random order (a cross-over design) for a month each and their metabolism, cholesterol, and other markers of insulin resistance (peripheral and hepatic insulin sensitivity) and inflammation (cortisol, csRP) were measured and compared with pre-weight loss and compared to each other (i.e. subject’s metabolic rate and cholesterol on low-fat was compared with the same subject’s met. rate and cholesterol on low-carb, etc.).  The diets were: low-fat diet (60% carb, 20% fat, 20% protein), low-glycemic index (40% carb, 40% fat, 20% protein), and very low-carb diet (10% carb, 60% fat, 30% protein).

Commentary on this clinical trial:  for more information from smarter, more-informed people than me, read these:

An insulin-hypothesis skeptic: Dr. Stephan Guyenet 

Insulin-hypothesis believers: Gary Taubes and Dr. Peter Attia.

Somebody in the middle: Dr. Robert Lustig