Hacker News Reader

Comment by yodelshady | Hacker News Reader

I've recently come off a 10 week pause in any significant exercise due to injury (and still can't do many strength exercises, though I was already mostly cardio). I made zero effort to change my eating habits, since I still have a job, and doing it is much easier - debatably only possible - if not distracted by hunger. My weight went down and is skyrocketing now that normal service is resumed.

Maybe I break the laws of physics, maybe my brain is lying to me about how much I ate, or maybe my endocrine system is doing things to maintain a setpoint, like it does with temperature, hydrostatic balance, CO2 levels, Na/K balance, anything else I've missed?

Given the above, how the ** do you suggest that CICO works as a predictive model over long terms? I don't know the first number, really don't know the second, and the second changes in response to the first.

View on HN

CICO is a very simple model that doesn't account for a lot of complexity in the human body and its nutrition.

Consider the following: 1500 calories worth of Snickers bar would not have the same effects as 1500 calories worth of beans and legumes or 1500 calories of pure soda. Satiety, nutrient and calorie availability, hormones (insulin), all differ depending on what you're eating -- and how often and how much you're eating. Even the combination of macro nutrients will have different effects: for example, fats + simple carbs is way worse than fats + proteins + fiber.

And exercise, mood, sunshine also factor in.

In the end, of course, thermodynamics holds and if you eat 1000 calories (no matter what form they take) when you need at a minimum 1500 then 100% yoi will lose weight. But how fast, how easy or how healthy that would be is a different story depending on what you're eating.

> Consider the following: 1500 calories worth of Snickers bar would not have the same effects as 1500 calories worth of beans and legumes or 1500 calories of pure soda.

How sure of that are you? Mark Haub, Nutrition Professor, famously ate a diet made entirely of "convenience store food" like Twinkies and chips for 10wks at a caloric deficit.

Despite a diet based on pure junk food, all of his health metrics improved and he lost 27lbs[0].

As someone who has lost well over 100lbs and who struggles constantly to deal with their weight, I can say from my experience that how many calories you eat is practically the only thing that matters. I have tried all the fad diets over the years and not a one of them has managed to lose me any weight and perform any better without strict calorie counting.

[0] http://www.cnn.com/2010/HEALTH/11/08/twinkie.diet.professor/...

I already address this point in my original comment:

"In the end, of course, thermodynamics holds and if you eat 1000 calories (no matter what form they take) when you need at a minimum 1500 then 100% yoi will lose weight. But how fast, how easy or how healthy that would be is a different story depending on what you're eating."

Except that's what I'm addressing: there is evidence that it really doesn't matter what you are eating, as long as you eat less.

Yes, I agree with you, I said:

"if you eat 1000 calories (no matter what form they take) when you need at a minimum 1500 then 100% you will lose weight"

And then I added:

"how fast, how easy or how healthy that would be is a different story depending on what you're eating."

You can lose weight (or maintain weight) only drinking 1000 calories worth of soda everyday. But it will be harder, way harder, than losing weight by eating 1000 calories of baked cauliflower with some nuts and some parmesan.

> for 10wks at a caloric deficit.

A human lifespan is not 10 weeks. Did he do that for the next 50 years? Because that's what's necessary.

IIRC the average American gains weight equivalent to a 50 kCal / day surplus, or 2.5%. By that same logic, going by weight gain and activity, I've started eating 70% more over the last two weeks, and I can't agree that's what happened. I eat out rarely and my groceries bill and cart contents have not changed.

I might believe the claim in isolation but, if all of his health metrics improved - sorry, I'm going to need a replication. Over 40 weeks.

> However, he did admit to eating veggies around his kids.

Oh look.

CICO is one of these things that are technically true, but useless in practice. What does the O mean here, exactly? It depends on metabolism. Some (few) people struggle to gain weight, some people struggle to not gain weight, consuming the same amount of calories. The former burn far more calories at rest than the latter. Some studies even show that caloric restriction slows down metabolism. This a survival feature of the body. The benefit of exercise is that it may(!) increase your metabolic rate so that you burn more calories at rest, but that's not something that will show up in some brief interventional study like the ones presented here.

> What does the O mean here, exactly?

Doesn't matter. Ignore the O, focus on the I. If you want to weigh a certain amount, look up what the sedentary metabolic estimate is for that weight and consume that number that despite any additional exercise you are doing.

The hard part is dealing with being hungry all the time, and despite the promises of fad diets everywhere, there is no solution to that that I have found that doesn't involve stimulants.

Metabolic rates are both individual and to some degree genetic. You can't just "look it up". In my text, I specifically mention research[1] showing caloric restriction can lower metabolic rate, which means that more weight will be gained for the same amount of I. This is a fundamental issue that would explain why diets don't work long-term.

[1] https://www.health.harvard.edu/blog/exercise-metabolism-and-...

The dirty secret about "metabolism" is that, except for an EXTRAORDINARILY SMALL part of the population, the range of metabolic differences between individuals is approximately equivalent to a single candy bar's worth of calories.

If you have a "slow metabolism" eat one candy bar's less worth of calories and you'll be fine.

Nine out of ten humans will be within 10% (less than a candy bar) of what an online BMR calculator spits out based on their sex, age, weight, and height.

The tenth out of ten humans is the person in the article you linked to, or an elite athlete.

If you are one of those one out of ten humans, you are or should be under the care of a physician.

Everyone else can count calories and adjust based on their observations.

BMR has the same factoid-spouting prigs as BMI. Yeah, we get it, you're a powerlifter so BMI doesn't work with you. Shut up and let the 99.97% of the population for who it works use it.

Yeah right, counting calories works for "99.7% of the population" but then everyone struggles with weight loss, somehow.

Cutting out that candy bar may work on paper, but the body will adapt to that reduced intake.

I'll cite the article I linked:

"One show contestant lost 239 pounds and achieved a weight of 191 pounds, yet six years later, after regaining 100 pounds of that lost weight, had to consume an 800-calorie-per-day diet to maintain his weight."

Now, you may argue that this person is an extreme outlier, and that's certainly true. However, it illustrates the principle. Differences as small as 10% can make a big impact over time.

I guess the subject in question was not put in a prison cell where 800 kcal was handled through a controlled opening?

The 130kg person burning just 800kcal/day is something highly suspicious. I am not sure if would be enough to lie in coma in a +30C room with a severe thyroid hormones deficiency to get to that level.

There are examples from hundreds of concentration/POW camps, periods of famine: if such extreme BMR lowering would be a thing, we would see it long time ago.

> Yeah right, counting calories works for "99.7% of the population" but then everyone struggles with weight loss, somehow.

Yes, because being hungry sucks. Why this has to be explicitly pointed out to you, I don't know.

(1) start with a reasonable estimate of daily expenditure. (2) based on (1) predict the food intake to achieve the desired weight change. (3) measure food intake. (4) measure weight change. (5) update (1) based on (3) and (4). goto (2).

I explicitly stated a predictive model, which that isn't.

I can come up with any old crap as advice if, afterwards, I can tell the recipient "you need to update the figures until my predictions stop being wrong". That just means the figures I gave were wrong.

Reply on news.ycombinator.com