Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

People are bad at reporting what they eat. That's a problem for dietary research

science.org

Post reply on HN
Print view View original

431–440 of 470 posts

Re: People are bad at reporting what they eat. That's a problem for dietary research

#431

Post by patanegra »

vanviegen wrote: post #293
foxyv wrote: post #240

Earlier quoted context omitted.

It doesn't help that food manufacturers intentionally make it hard to measure nutrition from most of their foods. They play around with serving sizes to hide carbohydrates making you have to do math just to keep up. Sometimes they will round down on grams of macros after setting the serving size so they can claim it has zero sugar when it does in fact have tons of sugar. Tic-tacs are the worst about this. They claim…

In the EU, food manufacturers are required to label macronutrients (and salt) in mg/100mg or mg/100ml for fluids. Easy to compare, works great.

[dead]

Re: People are bad at reporting what they eat. That's a problem for dietary research

#432

Post by djtango »

Ekaros wrote: post #321
djtango wrote: post #311

Earlier quoted context omitted.

Could you elaborate on what you mean? What does methane have to do with this and what is the "out" column? Calorimetry is just measuring the heat transfer from combustion, usually by measuring the temperature change of a known quantity of water in the classical experiment. You perform versions of it in high school and undergrad Calories are just a unit of energy, and heat can be related back to energy (joules for peo…

Well some of the food we eat generates flatulence of which 7% can be methane. Meaning this leaves our system without burning. As such in calorimeter it would be unburned fuel. Meaning that some calories are not absorbed failing the calories in and out equation.

Yes absolutely. Not only that but plenty of our food passes through undigested whatsoever. In theory we can't control that so we can only measure calories ingested, not calories absorbed but it hopefully sheds light on the fact that this is more complicated than just a number

Re: People are bad at reporting what they eat. That's a problem for dietary research

#433

Post by rapjr9 »

wnorris510 wrote: post #130

I've done research in this space for many years at Google AI and now at SnapCalorie. The thing I find interesting is how confident people are in their ability to estimate portion size visually, and in truth how wrong they all are. We published in CVPR (top peer reviewed academic conference for computer vision) and people are on average off by 53% and even trained professionals are still off by 40%. Basically if you w…

I was part of a project that did some work in this area also, we developed a machine learning based wearable to detect chewing:

https://www.ah-lab.cs.dartmouth.edu/publications/detecting-e...

Re: People are bad at reporting what they eat. That's a problem for dietary research

#434

Post by interludead »

OscarDC wrote: post #306
gadders wrote: post #45

Earlier quoted context omitted.

How old are you?

Not him but because your answer surprised me I chose to reply: at 34 it is also something I always wondered. Becoming obese always seemed a little extreme to me and I fail to imagine how someone could reach that state without the accordingly extreme food-related habits - though maybe I'm just lucky to have the "right" metabolism and thus cannot relate. Though even if obesity was always linked to eating disorders, I u…

It's influenced by a range of factors beyond eating habits

Re: People are bad at reporting what they eat. That's a problem for dietary research

#435

Post by interludead »

TypingOutBugs wrote: post #6

I try calorie count with My Fitness Pal and holy shit it’s a lot of effort. Eat out and you’re screwed (estimated at best). When you include sauces and oils etc it’s really hard to be accurate in the best of times, and it’s just a pain to keep on top of. Best option is to avoid any so you don’t have to count. I imagine almost everyone will add bad data in a study at some point with the best of intentions.

That's part of why there's such a push for better methodologies

Re: People are bad at reporting what they eat. That's a problem for dietary research

#436

Post by ikr678 »

exhilaration wrote: post #418
barnabyjones wrote: post #416

Earlier quoted context omitted.

A2 milk not only exists but is very popular in Asia actually. https://en.wikipedia.org/wiki/The_a2_Milk_Company

What does A2 milk have to do with lactose? I don't see lactose mentioned in your link.

Its a reference to the previous comment's 'specially engineered cows' quip - these cows do exist and produce a milk that is easier to digest (but still contains lactose).

Re: People are bad at reporting what they eat. That's a problem for dietary research

#437

Post by elektrolite »

FriedrichN wrote: post #21
elektrolite wrote: post #11

Earlier quoted context omitted.

I think being consistently inaccurate helps. If you always get the same thing at a certain restaurant, you can start by giving your best estimate of the calories in that meal. Then if your average weight doesn't move in the direction you want you can adjust your target calories to compensate.

That probably doesn't work either unless they work in an automated fashion. Did the chef put two or three dashes (official SI unit) of this or that on your meal? A a "dash" or "splash" or "spritz" of certain things can easily mean 100-200 kcal. And if you deal with things like meat, maybe the cut you get today is more or less lean than what you got last week. I think tracking calories for a couple of weeks can be ver…

It works medium term for lots of people. Helped me get visible abs. But I do agree that tracking calories for the rest of my life sounds exhausting.

Re: People are bad at reporting what they eat. That's a problem for dietary research

#438

Post by steveBK123 »

ikr678 wrote: post #436
exhilaration wrote: post #418

Earlier quoted context omitted.

What does A2 milk have to do with lactose? I don't see lactose mentioned in your link.

Its a reference to the previous comment's 'specially engineered cows' quip - these cows do exist and produce a milk that is easier to digest (but still contains lactose).

Very cool thanks for sharing

Re: People are bad at reporting what they eat. That's a problem for dietary research

#439

Post by Yizahi »

yapyap wrote: post #263
UomoNeroNero wrote: post #200

Earlier quoted context omitted.

I have been diabetic for 20 years. I have tried every method, app, plan, and tool, including systems falsely marketed as "smart." No method works or delivers decent results except for using a scale and weighing ALL the ingredients. For a diabetic, eating "out" is always a roll of the dice. The "fun" feedback from post-meal blood sugar is always a reminder of how "eyeballing a plate" is utterly useless.

I’m rather unknowledgeable on diabetes, so here’s a question that may seem basic: does choosing healthier meal, a salad instead of sweet ribs, not suffice for a good blood sugar?

Salad is great for diabetics. The problem is everything else:). Like for instance I've discovered that 99% of all rice is extremely bad, even good pasta is bad, potatoes are poison, bread also bad, and the list goes on. Fruits are bad too.

Cooking at home can be managed, and still heavily limiting. Eating out is a nightmare. First of all there are no "diabetes" places in the similar style to "vegan". And eating in a restaurant with at least some diabetes friendly selection of dishes is hard. For example there are may be 4 soup dishes. But 3 of them or even all will have either potatoes or pasta as ingredient (and leaving out said ingredient makes for a very mall meal, because those are often added to compensate). Salad section - the same issue, too often they have sweet syrup added for flavor. Anything Asian has rice or noodles in large quantities (I often wonder what diabetics in Asia eat). Second course dishes like meat or fish also sometimes contain sweet "surprises". All in all it is very hard task to find something, in a big city even.

Re: People are bad at reporting what they eat. That's a problem for dietary research

#440

Post by Fricken »

sethhochberg wrote: post #325
Fricken wrote: post #315

The thing is we're all experts at eating food, we've all been doing it our whole lives. You'd think in that time one would have cultivated an intuition about whether they are eating too much or too little regardless of the nutrition information.

Constant and plentiful supplies of food are, on an evolutionary scale, a somewhat recent thing. Agriculture itself is what, around 12k years old - a few hundred generations? We're all experts at eating food but having an instinctual understanding of exactly when to stop eating was probably much less important historically.

Cavemen could tell the difference between feast and famine and so can we. It doesn't mean we can control ourselves. I don't believe the crux of the problem is related to the information on the nutrition label, or how accurately we read it.

My theory is counting calories offers a convenient apparatus with which to perform some mental gymnastics when rationalizing why one is not actually eating too much when deep down one knows they eating too much.

All the emotional reasons for eating too much are messy and we would rather not deal with them. The connection between mood and gut can be quite strong, and can vary wildly from one person to another.

Post reply on HN