Sorta right. In the end there’s some sort of caloric restriction required to effectively lose weight. So, yes. All things being equal calories do count. The problem is all things are not equal.
Not all food macros are created the same. For instance, none of the three macros produce the same hormonal responses. Insulin gets secreted by eating carbs and protein. In the case of protein, the insulin is counterbalanced by glycogon. Not so in the case of carbs. Fat gets stored very efficiently, carbs less so. Overfeeding studies show that protein doesn’t get stored as fat much at all.
The different parts of your body charged with dealing with excess of a particular macro operate in different manners depending upon how full they are. The liver stores carbohydrates as glycogen until the glycogen stores get full. The liver then stores excess carbohydrates as fat in the liver until the liver gets full. The extra fat from the then liver overflows into the blood stream and you end up with high triglycerides, etc. Your fat cells are made to take a flux in and out of fat every day. Eat too often and that outflux doesn’t keep up with the influx. Do that too long and you become obese. It’s all a viscous circle that needs to be broken.
The killer combination (and I mean killer literally) is fat and carbs together. There are people (like the Potato diet guy) who can eat carbs alone without fat and lose weight. There are people who can eat a lot of fat and very low carbs and lose weight.
Eating both Low Carb and Low Fat along with higher protein is the winning combination for rapid fat loss. Study after study has shown this to be the best way to change your body composition (lose fat and retain lean mass or maybe even gain a little bit of lean mass).
I don’t care who you are if you eat around 20 grams of carbs or less, around 20 grams of fat, and maybe 150 gram of protein you will lose a significant amount of weight. You might be miserable doing it, or like me, you may enjoy seeing the weight fall off quickly. I really did enjoy it a lot. It’s around 800 calories and assuming you have a fair amount of body fat you can tolerate eating like that for quite some time. I don’t recommend this unless you have a fair amount of body fat to deal with.
The really strange thing about this is that the people who oppose this rapid fat loss method seem to prefer water fasting where they eat nothing for days to a diet where they eat 800 calories a day. It boggles the mind how not eating is easier than actually eating. Perhaps they were on low fat diets in the past and have some bad association from the hunger pains that come from low fat and high carb diets? There’s no blood sugar roller coaster with this method of losing weight since it is low carb.
For most keto folks it’s nothing more than cutting out the butter or other extra fats that they add to their food.
In theory, combining my Protein, Fat, and Carb results should yield a good set of maintenance macros. To do this I took the data and fit it to linear trendlines. The y-axis is weight change the day following the macro. Solving each of the equations for the Y-axis intercept gives the grams that result in the change from weight loss to gain the following day.
Re-running the trendline data as linear gives the following graphs.
Y Intercept Values
- Protein = 1.3802/.0066 = 209g = 837 kcal
- Fat = 1.8728/0.0144 = 164g = 1478 kcal
- Carbs = 0.7587/0.027 = 28 = 112 kcal
- Total Calories = 2427 kcal
This compares fairly well to the total calories Y-axis intercept value of 2.5254/0.0011 = 2296 kcal (about 5% difference). My median consumption was 2231 kcal.
Averages Across 150 Days
My averages across the 150 days is:
- Protein = 193g
- Fat = 155g
- Carbs = 24g
- Total = 2462 kcal
My conclusion is that my current maintenance numbers are pretty close.
Interesting study that looked at the energy from nuts (David J Baer Sarah K Gebauer Janet A Novotny. Walnuts Consumed by Healthy Adults Provide Less Available Energy than Predicted by the Atwater Factors. The Journal of Nutrition, Volume 146, Issue 1, 1 January 2016, Pages 9–13).
Results: One 28-g serving of walnuts contained 146 kcal (5.22 kcal/g), 39 kcal/serving less than the calculated value of 185 kcal/serving (6.61 kcal/g). The ME of the walnuts was 21% less than that predicted by the Atwater factors (P < 0.0001).
Conclusion: Consistent with other tree nuts, Atwater factors overestimate the metabolizable energy value of walnuts. These results could help explain the observations that consumers of nuts do not gain excessive weight and could improve the accuracy of food labeling.
This may explain why I weigh nuts and don’t seem to get the weight increase that the raw calories would indicate.
Similar results for another type of nut ( Baer DJ, Gebauer SK, Novotny JA. Measured energy value of pistachios in the human diet. Br J Nutr. 2012 Jan;107(1):120-5.). In that case the difference is less at about 5%.
Don’t take this as license to eat big bags of nuts. It’s only 20% for walnuts.
I did the same sort of analysis of my last 150 days of Cronometer data that I did for protein and carbs, but compared day-to-day weight change to the amount of fat consumed. Here’s the scatter diagram for that data:
For me, this explodes the myth that to lose more weight you need to eat more fat. To be more accurate it is true from around 60 to 120 grams of fat that the more fat I ate the more weight I lost the following day. But the curve quickly reverses after that and shows a fairly straight rise. Past around 175 g of fat the weigh increases the next day.
Since I’m looking at data, what is the effect of calories one day on weight the next day? Here’s the chart of calories vs weight change the next day.
Some observations about the data.
- The vertical axis is weight change the day after.
- The horizontal axis is calories eaten the day before the change.
- Calories in and weight increase out seems to work.
- My crossover point seems to be around 2450 calories. That’s pretty close to the Cronometer estimate of what I am expending.
- Next day increases are exaggerated versions of calories consumed. I went out to Red Robin last night and had three big plates of the salad bar. This morning I went up 3+ lbs. There’s no way I ate 10,000 calories at the salad bar, but I believe I could have 3 lbs of veggies in my system.
This confirms calories in and calories out on some rough level. It’s not simple and the error bars of uncertainty are huge, but taken together it shows that if I eat more the the 2450 calories I gain weight and if I eat less I lose weight.
One thing I’ve often wondered about since starting this low carb journey is just what my personal sensitivity to net carbs is. In other words, how many grams of net carbs can I eat without gaining weight? I’ve now got a 150 day long data set which gives me an answer. This chart shows how much the grams of carbs I eat one day affects my weight the next day.
About the chart:
- The vertical axis is weight gained the next day.
- The horizontal axis is grams of carbs.
- The red line on the chart is a 3rd order polynomial trendline as determined by EXCEL.
At less than 20 grams of carbs the line is pretty flat. It doesn’t seem like reducing my carbohydrates below 20 grams has much of a payoff in weight loss.
The turnaround point seems to be around 32 grams of net carbs. Past that point my weight gets progressively higher the next day.
From this I would conclude that my personal carbohydrate threshold is around 32 net grams of carbohydrates but that I can tolerate up to around 40 net grams without a large effect.
In my last post I looked at the question of whether eating protein one day causes weight gain the next day (“Protein Makes Me Gain Weight The Next Day”). Turned out protein didn’t make me gain weight all the way up to past 1 grams per pound of body weight.
This post looks at the maintenance macros I got from Ted Naiman’s website as described here (Maintenance Macros – Dr Ted Naiman). I have 150 days of Cronometer data and looked at the Protein to Non-Protein Energy ratio vs day-to-day weight gain.
Ted’s method sets the grams of protein starts to your body weight in lbs. Use the same number of grams of non-protein (carbs and fat) to maintain your weight. For me, that’s 170g of protein, 150g of fat, and 20g of carbs. With this method there’s no need to convert the number of grams to calories.
What does my own data show for Protein to non-protein ratio vs day-to-day ratio?
Here’s what I find interesting in the data:
- The vertical axis is weight change day-to-day. The zero line means that I did not change weight the next day.
- The day-to-day variance is pretty big.
- The horizontal axis is the ratio of Protein to Non-Protein grams. Most of my numbers were 1.0 or greater since I was trying to cut weight most of the time. That means I had more grams of protein than grams of fat+carbs.
- This graph looks at trendlines to generalize out the large day-to-day variances.
- The trendline is a 3rd order polynomial.
- There are only a couple of data points above a ratio of 2.0 to base the swing up after 2.25 on. It may not be a real swing.
- The best ratio for me to have the most loss the next day is around 1.75 grams of protein to grams of fat+carbs. Since I tend to hit my body weight or more in protein that would mean that I am eating relatively low fat on those days.
- There is a near zero crossing around 0.8:1 (which is more fat than protein). Ted’s numbers indicate the zero cross is around 1:1 which isn’t too far off from my numbers.
- Below the 0.8 I have a tendency to increase in weight.
All of this matches Ted’s concept. Here is Ted Naiman’s infographic:
Ted’s numbers for maintenance are pretty close to my actual measurements. I may have to eat a bit more fat to get to my goals of maintenance – which fits my experiences in the past month or so.
[This post replaces an earlier post which had an error in the spreadsheet data].
A common complaint I see on the Internet is
“When I eat more protein my weight goes up the next day.”
I rarely see anyone dig deeper into the question of whether higher consumption of protein actually causes greater weight gain the next day spread across time. A single day doesn’t say much about the effect of protein on weight.
A lot of days of eating protein
I have the last 150 days of data from Cronometer so I looked at how much my weight goes up the day after I eat protein. This is a chart of protein consumption compared to weight change on the following day.
A couple of things to note about this graph:
- The diagram is a scatter diagram.
- The Y axis is weight change between two adjacent days. Day to day weight fluctuations are all over the place.
- The X axis is grams of protein eaten the day before the weight change.
- The data is all over the place.
- The data appears to increase somewhat. More protein should on some level correspond to more weight gain (or less weight loss) since it’s more calories.
- The red line is the trendline that EXCEL determined.
- The trendline is a 3rd order polynomial.
- The deviation from the trendline is really big.
- The trendline shows that under 190 grams of protein (a whole lot of protein) results in weight loss (on the average).
- My weight is currently around 167 lbs.
- About 190 grams of protein results in weight stability (a bit over 1 gram of protein per lb of my weight).
- More protein than 190 grams results in a very small weight increases all the way to over 30 grams of protein.
- Very low protein (less than 100 grams) results in the most weight lost. It is not clear what the nature of this loss is from this chart. Is it more lean body mass lost? Is it just the low calorie amount for the day?
- I don’t have any very low protein data since I think it’s a serious mistake to undereat protein.
Protein, up more than 1 gram per lb of body weight did not result in average day-to-day weight gain.
Your mileage may vary. I am not sure how it could vary by much, but it may.
Apple carts are turned over. Money changers have their tables tossed. The world is changed by contrarians. I don’t always or even often agree with the contrarians but they do make me think. And sometimes they interact with you.
Low carb, in at least one sense, is all about being contrarian. Or at least it was back in the day of Dr Atkins.
There’s a particular form of contrarian that I like – those who are evidence based. Here’s a few of these contrarians.
Hormones Demystified – An Endocrinologist who is definitely a contrarian. Not too sure about Low Carb but at least gets it. He has mad respect for this crop of Engineers who are attacking medical problems. His analysis of Joseph Kraft’s database of Insulin Resistance in thousands of patients is well worth reading.
Carb Sanity – Evelyn has her story and she’s sticking to it. You are sick because you eat too much. It’s all about energy balance and any evidence to the contrary gets ignored.
Free The Animal – Richard is no fan of Jimmy Moore and many of the low carb luminaries. He thinks they are just plain wrong on eating all of that extra fat in their diet. Thinks they are missing out on eating the fat off their bodies. And he’s 100% correct. He’s currently obsessed about bitcoin but still posts on Low Carb occasionally when events warrant.
If you have more, let me know.
It’s often claimed that a fat weighs more than muscle. Some even go so far as to claim that a pound of fat weighs more than a pound of muscle. There are even graphics produced which show huge differences in volume such as:
Or the much more ridiculous graphic:
What people really mean is that fat is less dense than muscle. Oil, for instance, floats on top of water. That is because oil is less dense than water.
On average, the density of fat is 0.9 g/mL. The density of muscle is 1.1 g/mL. Using the averages, 1 liter of muscle weighs 1.06 kg, or 2.3 lbs., while 1 liter of fat weighs .9 kg, or 1.98 lbs. (Source)
The difference is less than 20%.
(Image from Bannister Nutrition).