9 ms·

Show HN: Glycemic load calculator

68 points by assafmo 8y ago

entropie 8y ago
> Glycemic index (GI) > GI is a measure of the effects of a foods carbohydrate component on blood sugar levels. Carbohydrates that break down quickly during digestion and rapidly release glucose into the bloodstream are characterized as having a high GI value, whereas carbohydrates that are broken down slowly and release glucose into the bloodstream in a more gradual manner are characterized as having a low GI value. > Glycemic load (GL) > Glycemic Load takes into account a foods GI value and a standardized 100g portion size and is calculated as GL = GI x available CHO in a 100g serving / 100. Therefore the GL takes into account the amount of carbohydrate consumed and is a more accurate measure of the impact of a food on blood sugars. As a general rule foods that have a low GL usually have a low GI and those with a medium to high GL value almost always have a very high GI value. http://www.diogenes-eu.org/GI-Database/Default.htm http://www.diogenes-eu.org/GI-Database/Default.htm
- tommaho 8y ago
  Something I rarely see addressed as an outsider and occasional consumer of this kind of info: I like bagels - this tool reports a GI / GL of 72 / 42 for 100g. I like peanut butter - 27 and 2.7 for 100g of creamy unsalted. What is the GI / GL of a peanut-buttered bagel? Is it an average? Can I calculate it? I don't see much advice around working toward an average glycemic load for a meal versus avoiding specific evils. Is that line of thought flawed?
  - entropie 8y ago
    A portion or meal differs for like anyone. When I make a slice of bread its like twice the size the one my SO makes. That goes for everything. So you have to calculate it for yourself. I dont have much knowledge about this kind of stuff; i "researched" what GI and GL actually means and pasted it for others also curious.
  - lolc 8y ago
    In the end your metabolism will have its own way of digesting which may or may not be close to the very general glycemic load. Which is compounded by your metabolism reacting differently based on time of day, what you ate before, what after, whether you're sick or just in a different state. With that out of the way, of course it depends which things you eat together. Potatoes are high on the index but if you combine them with cheese (mmmm cheese) it will slow down digestion and thus lower the glycemic load. Similarly, bread has a high load. But add peanut butter and you slow it down. As a general rule: Yes, add fat to slow digestion. There may be methods to calculate the load of a compund meal but don't count too much on them. The glycemic index is an average of how people react to foods. It's already impractical to study all the meals in that fashion and still you wouldn't know how you individually reacted to them. Edit: If you want to watch how your body reacts, you can get a set of Freestyle Libre sensors from Abott which allows you to monitor your blood glucose levels. They're not super accurate, but they easily allow you to track up/down movements which is enough to see how long it takes you to digest a meal.
  - jmulho 8y ago
    I don't know how to calculate it. But I do know that the combination matters. Try this experiment. Eat a bagel plain for breakfast. Wait as long as you can for lunch. Note when you start feeling like you are starving and your hands get shaky (blood sugar crash). Next day eat the bagel with peanut butter. Note how you feel after the same amount of time. I predict the blood sugar crash will later, or won't happen at all. I have found that the secret to sustaining a vegan diet is beans (resistant starch). They release sugar into your blood over many hours, blunting the blood sugar roller coaster that would otherwise be caused by eating a lot of fruits and vegetables without a lot of meats and fat. I have done a version of the above proposed experiment with a foot long Subway sandwich, veggies only (no meat, no cheese, no mayo). I feel like I am starving in 3 hours. Add hummus, I feel fine for 6 hours. I know it adds some calories, but I'm convinced it is more about the resistant starch. According to this video https://nutritionfacts.org/video/beans-and-the-second-meal-effect/ https://nutritionfacts.org/video/beans-and-the-second-meal-e... eating beans in one meal even reduces the blood sugar spike in later meals! Another interesting thing. There is a Ted Talk about how glycemic response varies by individual due to different makeup of gut bacteria. I think it is this one: https://www.youtube.com/watch?v=0z03xkwFbw4 https://www.youtube.com/watch?v=0z03xkwFbw4. The guy claims he can analyze your gut bacteria and tell you which foods will spike your blood sugar. Subway might be bad for me but fine for you!
    - nradov 8y ago
      I've actually tried that peanut butter bagel experiment myself and didn't notice any significant difference. Of course experiments with a population of 1 and no control group are meaningless.
cranjice 8y ago
So I tried entering the term "pasta" and nothing happened. At first I thought it was broken, but then I entered pizza and there were results. An action button and some status detail (# of results found, etc.) might help the UI.
- assafmo 8y ago
  Weird, pasta is in the database. What happened to you is probably the slowness of the search. The database is loaded entirely to the client and searches are done in memory. Right now it's pretty jumpy even on high-end hardware. I need to fix this.
mygo 8y ago
you only designed and implemented for the happy path. when a user submits something that isn’t in your database absolutely nothing happens. a submit button would be a start, as well as some sort of “no results found”. Feedback is important to the user.
- assafmo 8y ago
  Thanks for your feedback! I'll address this.
  - dayjah 8y ago
    Also some form of typeahead prompt would be useful for a similar reason — for example: if one enters “bread”, being able to select the type of bread from a known list is helpful.
    - assafmo 8y ago
      It's pretty hard because of the source data. I'll see what I can do about dividing search results into food categories. Thanks for your feedback.
- assafmo 8y ago
  Done. :-)
harigov 8y ago
Are you sure this data is correct? It shows that boiled brown rice has a glycemic load higher than white rice.
- assafmo 8y ago
  The data was taken from http://www.diogenes-eu.org/GI-Database/Default.htm http://www.diogenes-eu.org/GI-Database/Default.htm (https://github.com/assafmo/gl-calc/tree/master/src_data https://github.com/assafmo/gl-calc/tree/master/src_data)
  - assafmo 8y ago
    Moved to https://github.com/assafmo/gl-calc/tree/master/data https://github.com/assafmo/gl-calc/tree/master/data
- magic_beans 8y ago
  There are a wide variety of rice types, and YES, some white rice varieties have a lower glycemic index than brown rice varieties.
  - jsmp 8y ago
    also boiled rice absorb more water, this weight includes that water
goldenkey 8y ago
A little off-topic but I recently realized the cultural history of "appetizers." Most of us, myself included, are used to appetizers being a restaurant thing. But I did some digging and it appears there is a valid scientific use. "Whetting one's appetite" is useful for everyone if they want to eat healthier meals. The average person consumes _pounds_ of food daily. If you're like me, it is difficult to eat a large meal even if it is tasty and healthy only if by the size of it. A good meal would include an appetizer to get one acclimated to the larger, less tasty, more work-required meal that is next. Realizing this helped me adjust all my meals to include some kind of appetizer. Now I eat more healthy food and it isnt as much of a mental struggle. Hope this realization helps someone else too..
- lolc 8y ago
  I don't understand your comment. Where I live people generally eat way too much. Me included. It's common understanding here that healthy meals would be smaller but people still eat big meals. If I strain I can understand your comment to say that appetizers allow us to eat the healthier food that follows. This would require the second course to be more healthy than the first. In my experience this is a rare occurrence.
- magic_beans 8y ago
  You are a rare case. Most people love eating. Most people do not struggle to eat large amounts food. Most people actively enjoy eating large amounts of food. No other internet forum on the world is as full of people who hate eating as HN and thinspo-gram.
ruc0la 8y ago
I was wondering: can you really take out from the equation the bodyweight/fit level/age variables? I'm heavy and also very tall and it seems to me an over simplified service using only a macro sheet of products.
- ucaetano 8y ago
  No. And even if you take in account those factors, the resulting number is irrelevant on an individual basis due to huge variation between individuals.
jl2718 8y ago
These numbers are completely meaningless for nutrition purposes, and should be generally ignored and replaced with much simpler metrics like CHO digestion half-life. Glucose levels are increased by digestion/conversion of CHO, and decreased by the effect of insulin. Different types of CHO digest slower(fiber), get into the blood stream faster (glucose) or release more insulin (fructose). A low glycemic load could mean either that all the sugar was immediately deposited in fat cells and now you’re hungry again because it’s dropping and your ghrehlin is spiking, or that your food is still digesting slowly. Do not be deceived by the low GI/GL. Fructose goes right into your fat cells and makes you hungry by spiking insulin. Seriously. Ignore GI/GL.
- woodandsteel 8y ago
  Question: is what you are saying widely accepted by the experts who study human digestion, metabolism, and appetite, or is it a minority view? On a more anecdotal level, I don't find that eating apples makes me hungry shortly afterwards.
  - robbiep 8y ago
    I studied medicine at the university that created and trademarked he glycaemic Index (GI). Prof Brand-Miller gave us lectures in physiology and nutrition. Even as she was giving them there was an understanding in the cohort that GI was flawed. for example, a high fructose meal is extremely low GI but that says nothing about either it’s nutritional value or how it will affect your bosy’s insulin response, which when you get down to it is the key focus of this flawed measure. The defence proffered by faculty (who had a vested interest in promoting it) was weak in our eyes and has only grown weaker over time
    - mygo 8y ago
      what do you recommend people use as a metric to predict insulin response instead?
      - robbiep 8y ago
magic_beans 8y ago
Glycemic load is DRAMATICALLY affected by what you are pairing your foods with. Add fat or fiber and the GL drops. This tool isn't all that helpful.
- mrfusion 8y ago
  Are you saying I could take a tablespoon of oil after a donut and it will change the glycemic load? I have been taking a cup of Metamucil after I have a dessert with that same theory in mind. Maybe it’s been helping.
mrfusion 8y ago
I wish there was something like this to see how foods would affect insulin levels. I’ve taken to calling insulin the fat storage hormone. Surprisingly glycemic index doesn’t always correspond to insulin level rise.
- DanielBMarkham 8y ago
  There is. I heard somebody is doing an insulinogenic index. Sorry, I don't have a link. I've been meaning to look into it but haven't had the time.
  - mrfusion 8y ago
    Here’s a start. And it’s intersting reading. https://optimisingnutrition.com/tag/insulinogenic-foods/ https://optimisingnutrition.com/tag/insulinogenic-foods/
- rbommisetti 8y ago
  My father has diabetes and I'm looking to create an app like this. Can you elaborate on your idea?
dzhiurgis 8y ago
Reminds my little toy project: https://aminosplore.metacursion.com/ https://aminosplore.metacursion.com/ It displays 7 main protein content and lets to compare if you wanna build whole picture. It's super alpha and barely works (Chrome desktop only, loads entire ~40mb SQLite-webassembly database...)
- assafmo 8y ago
  Cool!
contingencies 8y ago
Searched for walnuts. Got half a dozen pure walnut records (not walnut cake, etc.), GI was always 24 but GL ranged from 1-4. Source says GL is better indicator than GI. Data is all over the place and publication on which it is based is nearly ten years old. Unsure how to interpret. Inclined to ignore. PS. If anyone has a serious nutrition background and wants to develop algorithms for dietary value prediction based on customized meals and disparate data sources, we are hiring.
jdc 8y ago
Not sure where to find CHO digestion half-life figures
mstudio 8y ago
Very interesting! I can't speak to the accuracy of the data, but the overall app is quick, simple and easy to use. One minor UI suggestion: in the live search results, perhaps color code the GI, as you do on the results page. e.g. for "Apple", "GI:38" would be in Green to denote "low" instead of Teal.
- assafmo 8y ago
  Thanks for the feedback! I think I'll do this. :-)
- assafmo 8y ago
  Done.

68 points by assafmo 8y ago

> Glycemic index (GI) > GI is a measure of the effects of a foods carbohydrate component on blood sugar levels. Carbohydrates that break down quickly during digestion and rapidly release glucose into the bloodstream are characterized as having a high GI value, whereas carbohydrates that are broken down slowly and release glucose into the bloodstream in a more gradual manner are characterized as having a low GI value. > Glycemic load (GL) > Glycemic Load takes into account a foods GI value and a standardized 100g portion size and is calculated as GL = GI x available CHO in a 100g serving / 100. Therefore the GL takes into account the amount of carbohydrate consumed and is a more accurate measure of the impact of a food on blood sugars. As a general rule foods that have a low GL usually have a low GI and those with a medium to high GL value almost always have a very high GI value. http://www.diogenes-eu.org/GI-Database/Default.htm http://www.diogenes-eu.org/GI-Database/Default.htm

tommaho 8y ago
Something I rarely see addressed as an outsider and occasional consumer of this kind of info: I like bagels - this tool reports a GI / GL of 72 / 42 for 100g. I like peanut butter - 27 and 2.7 for 100g of creamy unsalted. What is the GI / GL of a peanut-buttered bagel? Is it an average? Can I calculate it? I don't see much advice around working toward an average glycemic load for a meal versus avoiding specific evils. Is that line of thought flawed?
- entropie 8y ago
  A portion or meal differs for like anyone. When I make a slice of bread its like twice the size the one my SO makes. That goes for everything. So you have to calculate it for yourself. I dont have much knowledge about this kind of stuff; i "researched" what GI and GL actually means and pasted it for others also curious.
- lolc 8y ago
  In the end your metabolism will have its own way of digesting which may or may not be close to the very general glycemic load. Which is compounded by your metabolism reacting differently based on time of day, what you ate before, what after, whether you're sick or just in a different state. With that out of the way, of course it depends which things you eat together. Potatoes are high on the index but if you combine them with cheese (mmmm cheese) it will slow down digestion and thus lower the glycemic load. Similarly, bread has a high load. But add peanut butter and you slow it down. As a general rule: Yes, add fat to slow digestion. There may be methods to calculate the load of a compund meal but don't count too much on them. The glycemic index is an average of how people react to foods. It's already impractical to study all the meals in that fashion and still you wouldn't know how you individually reacted to them. Edit: If you want to watch how your body reacts, you can get a set of Freestyle Libre sensors from Abott which allows you to monitor your blood glucose levels. They're not super accurate, but they easily allow you to track up/down movements which is enough to see how long it takes you to digest a meal.
- jmulho 8y ago
  I don't know how to calculate it. But I do know that the combination matters. Try this experiment. Eat a bagel plain for breakfast. Wait as long as you can for lunch. Note when you start feeling like you are starving and your hands get shaky (blood sugar crash). Next day eat the bagel with peanut butter. Note how you feel after the same amount of time. I predict the blood sugar crash will later, or won't happen at all. I have found that the secret to sustaining a vegan diet is beans (resistant starch). They release sugar into your blood over many hours, blunting the blood sugar roller coaster that would otherwise be caused by eating a lot of fruits and vegetables without a lot of meats and fat. I have done a version of the above proposed experiment with a foot long Subway sandwich, veggies only (no meat, no cheese, no mayo). I feel like I am starving in 3 hours. Add hummus, I feel fine for 6 hours. I know it adds some calories, but I'm convinced it is more about the resistant starch. According to this video https://nutritionfacts.org/video/beans-and-the-second-meal-effect/ https://nutritionfacts.org/video/beans-and-the-second-meal-e... eating beans in one meal even reduces the blood sugar spike in later meals! Another interesting thing. There is a Ted Talk about how glycemic response varies by individual due to different makeup of gut bacteria. I think it is this one: https://www.youtube.com/watch?v=0z03xkwFbw4 https://www.youtube.com/watch?v=0z03xkwFbw4. The guy claims he can analyze your gut bacteria and tell you which foods will spike your blood sugar. Subway might be bad for me but fine for you!
  - nradov 8y ago
    I've actually tried that peanut butter bagel experiment myself and didn't notice any significant difference. Of course experiments with a population of 1 and no control group are meaningless.

cranjice 8y ago

So I tried entering the term "pasta" and nothing happened. At first I thought it was broken, but then I entered pizza and there were results. An action button and some status detail (# of results found, etc.) might help the UI.

assafmo 8y ago
Weird, pasta is in the database. What happened to you is probably the slowness of the search. The database is loaded entirely to the client and searches are done in memory. Right now it's pretty jumpy even on high-end hardware. I need to fix this.

mygo 8y ago

you only designed and implemented for the happy path. when a user submits something that isn’t in your database absolutely nothing happens. a submit button would be a start, as well as some sort of “no results found”. Feedback is important to the user.

assafmo 8y ago
Thanks for your feedback! I'll address this.
- dayjah 8y ago
  Also some form of typeahead prompt would be useful for a similar reason — for example: if one enters “bread”, being able to select the type of bread from a known list is helpful.
  - assafmo 8y ago
    It's pretty hard because of the source data. I'll see what I can do about dividing search results into food categories. Thanks for your feedback.
assafmo 8y ago
Done. :-)

harigov 8y ago

Are you sure this data is correct? It shows that boiled brown rice has a glycemic load higher than white rice.

assafmo 8y ago
The data was taken from http://www.diogenes-eu.org/GI-Database/Default.htm http://www.diogenes-eu.org/GI-Database/Default.htm (https://github.com/assafmo/gl-calc/tree/master/src_data https://github.com/assafmo/gl-calc/tree/master/src_data)
- assafmo 8y ago
  Moved to https://github.com/assafmo/gl-calc/tree/master/data https://github.com/assafmo/gl-calc/tree/master/data
magic_beans 8y ago
There are a wide variety of rice types, and YES, some white rice varieties have a lower glycemic index than brown rice varieties.
- jsmp 8y ago
  also boiled rice absorb more water, this weight includes that water

goldenkey 8y ago

A little off-topic but I recently realized the cultural history of "appetizers." Most of us, myself included, are used to appetizers being a restaurant thing. But I did some digging and it appears there is a valid scientific use. "Whetting one's appetite" is useful for everyone if they want to eat healthier meals. The average person consumes _pounds_ of food daily. If you're like me, it is difficult to eat a large meal even if it is tasty and healthy only if by the size of it. A good meal would include an appetizer to get one acclimated to the larger, less tasty, more work-required meal that is next. Realizing this helped me adjust all my meals to include some kind of appetizer. Now I eat more healthy food and it isnt as much of a mental struggle. Hope this realization helps someone else too..

lolc 8y ago
I don't understand your comment. Where I live people generally eat way too much. Me included. It's common understanding here that healthy meals would be smaller but people still eat big meals. If I strain I can understand your comment to say that appetizers allow us to eat the healthier food that follows. This would require the second course to be more healthy than the first. In my experience this is a rare occurrence.
magic_beans 8y ago
You are a rare case. Most people love eating. Most people do not struggle to eat large amounts food. Most people actively enjoy eating large amounts of food. No other internet forum on the world is as full of people who hate eating as HN and thinspo-gram.

ruc0la 8y ago

I was wondering: can you really take out from the equation the bodyweight/fit level/age variables? I'm heavy and also very tall and it seems to me an over simplified service using only a macro sheet of products.

ucaetano 8y ago
No. And even if you take in account those factors, the resulting number is irrelevant on an individual basis due to huge variation between individuals.

jl2718 8y ago

These numbers are completely meaningless for nutrition purposes, and should be generally ignored and replaced with much simpler metrics like CHO digestion half-life. Glucose levels are increased by digestion/conversion of CHO, and decreased by the effect of insulin. Different types of CHO digest slower(fiber), get into the blood stream faster (glucose) or release more insulin (fructose). A low glycemic load could mean either that all the sugar was immediately deposited in fat cells and now you’re hungry again because it’s dropping and your ghrehlin is spiking, or that your food is still digesting slowly. Do not be deceived by the low GI/GL. Fructose goes right into your fat cells and makes you hungry by spiking insulin. Seriously. Ignore GI/GL.

woodandsteel 8y ago
Question: is what you are saying widely accepted by the experts who study human digestion, metabolism, and appetite, or is it a minority view? On a more anecdotal level, I don't find that eating apples makes me hungry shortly afterwards.
- robbiep 8y ago
  I studied medicine at the university that created and trademarked he glycaemic Index (GI). Prof Brand-Miller gave us lectures in physiology and nutrition. Even as she was giving them there was an understanding in the cohort that GI was flawed. for example, a high fructose meal is extremely low GI but that says nothing about either it’s nutritional value or how it will affect your bosy’s insulin response, which when you get down to it is the key focus of this flawed measure. The defence proffered by faculty (who had a vested interest in promoting it) was weak in our eyes and has only grown weaker over time
  - mygo 8y ago
    what do you recommend people use as a metric to predict insulin response instead?
    - robbiep 8y ago

magic_beans 8y ago

Glycemic load is DRAMATICALLY affected by what you are pairing your foods with. Add fat or fiber and the GL drops. This tool isn't all that helpful.

mrfusion 8y ago
Are you saying I could take a tablespoon of oil after a donut and it will change the glycemic load? I have been taking a cup of Metamucil after I have a dessert with that same theory in mind. Maybe it’s been helping.

mrfusion 8y ago

I wish there was something like this to see how foods would affect insulin levels. I’ve taken to calling insulin the fat storage hormone. Surprisingly glycemic index doesn’t always correspond to insulin level rise.

DanielBMarkham 8y ago
There is. I heard somebody is doing an insulinogenic index. Sorry, I don't have a link. I've been meaning to look into it but haven't had the time.
- mrfusion 8y ago
  Here’s a start. And it’s intersting reading. https://optimisingnutrition.com/tag/insulinogenic-foods/ https://optimisingnutrition.com/tag/insulinogenic-foods/
rbommisetti 8y ago
My father has diabetes and I'm looking to create an app like this. Can you elaborate on your idea?

dzhiurgis 8y ago

Reminds my little toy project: https://aminosplore.metacursion.com/ https://aminosplore.metacursion.com/ It displays 7 main protein content and lets to compare if you wanna build whole picture. It's super alpha and barely works (Chrome desktop only, loads entire ~40mb SQLite-webassembly database...)

assafmo 8y ago
Cool!

contingencies 8y ago

Searched for walnuts. Got half a dozen pure walnut records (not walnut cake, etc.), GI was always 24 but GL ranged from 1-4. Source says GL is better indicator than GI. Data is all over the place and publication on which it is based is nearly ten years old. Unsure how to interpret. Inclined to ignore. PS. If anyone has a serious nutrition background and wants to develop algorithms for dietary value prediction based on customized meals and disparate data sources, we are hiring.

jdc 8y ago

Not sure where to find CHO digestion half-life figures

mstudio 8y ago

Very interesting! I can't speak to the accuracy of the data, but the overall app is quick, simple and easy to use. One minor UI suggestion: in the live search results, perhaps color code the GI, as you do on the results page. e.g. for "Apple", "GI:38" would be in Green to denote "low" instead of Teal.

assafmo 8y ago
Thanks for the feedback! I think I'll do this. :-)
assafmo 8y ago
Done.