Have you ever wondered whether it is okay to eat bananas, grapes, or carrots? These are foods women are often told to avoid because of their “high” sugar or carbohydrate content, a fear usually traced back to something called the glycemic index.

Fortunately, that fear is usually unfounded. In this post, I want to walk you through what the glycemic index actually measures, why the glycemic load is a more useful tool, and why even that does not capture the full picture. My goal is that, by the end, you will have a framework that is simpler to use (and more accurate) than any lookup table.

This is the companion to my recent post on the two levers that support healthy blood sugar. If you have not read that one yet, it is worth a look, because everything here points back to that approach.

The Fear That Started It All (A Short History Lesson)

In the 1980s, researchers set out to determine how quickly dietary carbohydrates converted to blood glucose, as a way to manage the insulin response in people with diabetes. The initial study was done on 62 commonly eaten foods, and out of it the glycemic index was born.

The intention was to provide a tool to those whose blood sugar regulation now required significant intervention. But as with many nutrition tools that start off with good intentions, the actual impact has been very different.

Panic seemed to take hold about fruits and vegetables that “have so much sugar in them” when the glycemic index entered popular conversation, and we have been dealing with the fallout of that partial story ever since. Even the creation of the glycemic load by Harvard researchers in the late 1990s, which was meant to make the glycemic index information more applicable, hasn’t been able to undo the damage. Nor has the research that continually shows that the GI of a food or diet cannot be used to predict disease risk or health outcomes.

The biggest problem? The glycemic index was never designed to answer the question most of us are actually asking, which is simply: how will this food, eaten the way I normally eat it, affect my blood sugar?

I hope that, once you understand what the index really measures, you will stop fearing whole, nourishing foods and start making decisions with more confidence and ease.

What the Glycemic Index Actually Measures

As mentioned above, the glycemic index (GI) was originally developed as a tool for managing diabetes. Each food is given a rank from 0 to 100, and here is how that rank is determined.

A group of people is recruited for the study and asked to fast overnight. The next morning, they eat a measured portion of a single test food, and nothing else. The amount they eat is whatever quantity of that food contains 50 grams of digestible carbohydrate. Their blood glucose is then tracked for the two hours following the meal and plotted on a graph.

On a different day, the same people return, again fasted, and eat 50 grams of pure glucose. Their blood sugar is tracked the same way. Pure glucose is assigned the reference value of 100, and the test food is given a number relative to it, based on how the participants’ blood glucose responded.

Take a moment to notice how controlled those conditions are. A single food, eaten alone, on a completely empty stomach, in a quantity calculated to deliver exactly 50 grams of carbohydrate. Yes, we might fast overnight. And yes, on occasion we might eat just one food for breakfast. But most of us, most of the time, do not eat a single food in isolation, and possibly not enough of the food to provide 50 grams of available carbohydrate. That gap between the test and real life is where the trouble lies.

Where the Glycemic Index Breaks Down

The 50-gram-of-carbohydrate requirement is where the index becomes horribly misleading for carbohydrate-containing foods that are not concentrated sources of carbohydrate.

Take grapes, as an example. Grapes have a glycemic index of about 53, which sits just below the “moderate” range in the index. But to actually consume 50 grams of carbohydrate from grapes, you would need to eat roughly 190 grams of them, about 7 ounces, or a very large bunch. If you are reaching for grapes as a snack, you are probably eating a handful, less than half the amount that was measured to generate the GI number.

Carrots make the point even more dramatically. To eat 50 grams of digestible carbohydrate from carrots, you would have to eat just over a pound of them. I do not know anyone sitting down to a pound of carrots in one go.

A banana’s glycemic index can be as high as 62, but you’d need to eat ~2 bananas that were each 7-8 inches long.

To be clear, it is not that the glycemic index is lying about the impact a food can have on blood glucose levels. Rather, it is answering a question that simply is not relevant for these foods: how much does blood glucose rise when someone eats an unrealistic amount of this food, and nothing else, on an empty stomach? That is not a question that helps you decide whether to enjoy a handful of baby carrots or to add a banana to your smoothie.

Glycemic Load: A Better Tool, But Still Incomplete

This is where the glycemic load (GL) comes in. It was created to correct the portion problem by factoring in a realistic serving size rather than an arbitrary 50 grams of carbohydrate.

The difference is striking. Grapes have a glycemic load of about 7, clearly a “low glycemic food”, compared to a glycemic index that flirts with the moderate range. Raw carrots have a glycemic load of about 2 for a normal serving (roughly what fits in a cupped hand), compared to a glycemic index that made them sound like something to be limited. And the banana’s glycemic index is enough to put fear into anyone who is trying to keep their blood sugar in a healthy range, while the glycemic load also makes it a “low glycemic food”.

Without question, the glycemic load tells a much more honest story about how these foods behave when eaten in their usual amounts. But the glycemic load has its own limitation.

It is still calculated from how a food behaves eaten in isolation, and in fact it is not measured directly at all. It is derived mathematically from the glycemic index. So while it is genuinely more useful for everyday decisions, it inherits some of the same blind spots as the number it comes from.

Six More Reasons Not to Rely on the Glycemic Index or Glycemic Load

Beyond the inherent problems mentioned above, there are six more reasons I do not rely on the glycemic index or glycemic load as my main guide, and they matter especially if you are a busy woman in midlife who does not need one more confusing system to manage.

1. The data is not consistent, between databases or even within them.

If you are trying to figure out if something is a low glycemic food, you will often find different numbers from different sources, and sometimes conflicting numbers within the same source. Raw carrots are a good example: published values range from the teens to the forties depending on the variety, how they were prepared, and the particular group of people tested. As a person with a math degree, this problem bothers me a lot, because the lack of agreement within the foundational data undermines the precision the numbers appear to offer.

2. Ripeness changes everything.

As fruit ripens, its starches convert to sugar, so the carbohydrate enters the bloodstream faster. A very ripe banana has a glycemic index of 62+, while an underripe one sits around 31. But when you grab a banana for your smoothie, you probably are not assessing exactly how ripe it is, and the lookup table cannot tell you either.

3. Preparation matters.

How you prepare a food shifts its glycemic index and load. Raw carrots come in around 24 (low glycemic food), while boiled carrots range between 32-60 (low-moderate glycemic food). And something interesting happens with starchy foods like rice and potatoes: cooking them and then cooling them for 12 to 24 hours before consumption increases their resistant starch, a type of carbohydrate that, like fiber, your body cannot digest or absorb. The glycemic index has no idea whether you plan to eat your rice hot tonight or cold tomorrow.

4. The form of the food matters.

Whole, fresh foods generally have a lower glycemic index than their processed versions. Fruit juice ranks higher than the whole fruit it came from, because the fiber has been stripped away. Dried fruit poses a similar problem. With the water removed, it is far easier to eat five or six dried peaches than it would be to sit down to five or six fresh ones. And consider rice: a cup of cooked brown rice and a cup of cooked white rice contain roughly the same total carbohydrate, but the brown rice still has its bran and germ, and therefore its fiber, so its carbohydrate enters the bloodstream more slowly. Same plant, different processing, different effect.

5. The serving size used in research may not be the serving size that is appropriate for you.

The glycemic index, and even the glycemic load, is built around a standardized portion: the glycemic index uses whatever amount delivers 50 grams of carbohydrate, and the glycemic load uses an assumed “typical” serving. But neither one knows you. The right portion of any food depends on your body, your activity, your hunger, your goals, and how that food fits into the rest of your day. A serving size that was considered reasonable by a scientist may be more or less than what serves you in a given moment, which means the number attached to that serving may not translate to your plate.

6. The whole meal changes the outcome, and this is the big one.

This is the most important limitation of all. Protein, fat, and fiber all slow the digestion and absorption of carbohydrate. So any carbohydrate-containing food eaten as part of, or right after, a balanced meal produces a very different blood glucose response than the same food eaten alone. The glycemic index measures food in isolation, which is almost never how we actually eat. That single fact limits the usefulness of these numbers more than anything else.

The Simpler Approach to Managing Blood Glucose: Back to the Two Levers

Here is the good news, and it connects directly to the two levers I wrote about last week. Remember those two levers for supporting healthy blood sugar are: 1) how much glucose enters your body at one time, and 2) how quickly it gets there.

The glycemic index and glycemic load are really just trying to estimate those two things (how much glucose a food delivers and how fast), but only accomplishes that goal for foods eaten in isolation. You can get to the same place far more simply, without ever opening a lookup table:

• Reduce the starchy, sweetened, and processed foods you eat, which lowers the total glucose load of any meal or snack.
• Build your meals and snacks with protein, fat, and/or fiber, which slows how quickly any glucose reaches your bloodstream.

That is the entire framework. Once you get the hang of recognizing which foods on your plate are starchy, sweetened, or processed, and how to round out a meal when carbohydrate-containing foods are part of it, you no longer need anyone’s glycemic index database. You will have something better: an understanding you can apply to any food, in any combination, anywhere.

Frequently Asked Questions

What is the difference between glycemic index and glycemic load?

The glycemic index (GI) measures how quickly a food raises blood sugar when you eat enough of it to deliver 50 grams of carbohydrate, eaten alone on an empty stomach. The glycemic load (GL) adjusts for a realistic serving size, so it gives a more honest sense of how a normal portion affects you. The key difference: GI ignores how much you actually eat, while GL accounts for it. That is why a banana’s glycemic index can be moderate-sounding but it still be a low glycemic food.

How do I find the glycemic index of a food?

You can look up glycemic index values in several public databases, the most established being the one maintained by the University of Sydney. You’ll also find glycemic load values listed alongside them in many sources. Here’s the catch worth considering before you start: the same food often carries different values across databases, and sometimes within a single one, because the number depends on the food’s ripeness, how it was prepared, and the specific people tested. So a lookup will give you a ballpark, not a precise answer for the food you are about to consume. If you find yourself reaching for a database often, that’s usually a sign that you would be better served by learning the simple framework instead: reduce the starchy, sweetened, and processed foods, and pair any carbohydrate with protein, fat, or fiber. That approach works for any food, in any combination, without a single lookup.

How do you calculate glycemic load?

The formula is straightforward: glycemic load equals the glycemic index of the food multiplied by the grams of carbohydrate in your serving, divided by 100. So a food with a glycemic index of 50 and 20 grams of carbohydrate in a serving has a glycemic load of 10. That’s it.

But notice what that formula reveals. Glycemic load isn’t independently measured, it’s calculated from the glycemic index, which means it carries forward all the same limitations: it still reflects a food eaten alone, in a standardized portion, with all the database inconsistency baked in. The math is simple, but it can’t tell you how that food will behave on your plate, alongside other foods, prepared your way. That’s why I’d rather you learn to recognize starchy, sweetened, and processed foods and build balanced meals than spend time running calculations.

Are carrots and grapes bad for blood sugar?

For most people, no. Both have been unfairly judged because of their glycemic index, but their glycemic load, which reflects a realistic serving, makes them both low glycemic foods. Carrots come in around 2 and grapes around 7, both firmly in the low range. They also come packaged with water, fiber, and nutrients that slow how their natural sugars are absorbed and provide additional health benefits beyond blood glucose levels. Eaten in normal amounts, and especially alongside protein or fat (since they have their own fiber), they are not foods most people need to fear.

Does popcorn spike blood sugar?

It depends! And the “it depends” is exactly the point of this post. Popcorn is a starchy food, so on its own and in a large amount it can certainly raise blood sugar more than desired, especially the kind drenched in sugar or made to be quick and ultra-processed. But air-popped popcorn in a moderate portion, eaten alongside something with fat, protein, or fiber, behaves very differently than a big tub on an empty stomach. This is your two levers in action: how much you eat at once, and how quickly that glucose gets in. Rather than asking whether popcorn is “good” or “bad,” a more useful question is how much you’re having and what you’re having it with. That same question works for any food, which is what makes it more powerful than any glycemic index value.

Conclusion

The glycemic index gave us a partial story, and partial stories tend to create more fear than clarity. It was never meant to tell you whether a banana belongs in your smoothie or whether carrots deserve a place on your plate. The glycemic load improved on it, but still cannot account for how you actually eat: in combinations, in varying ripeness and preparation, in real life.

So rather than chasing numbers that cannot agree with each other, you can lean on something steadier. Reduce the starchy, sweetened, and processed foods, and build balanced meals with protein, fat, and fiber. That is a framework you can carry with you, no lookup table required.

In my next post, I dig into sweeteners in all their forms, including the ones marketed as healthier alternatives, so you can decide if and how they fit into your life. I have mentioned more than once that sweetened foods affect your blood sugar, and they deserve a closer look. Stay tuned.

And if you would like personalized support in building an approach that fits your body and your life, especially as it changes in midlife, let’s chat. I’d love to support you as you figure out how to achieve and maintain those healthy blood sugar levels you desire.

Book your free 15-minute connect call.

Image by ifd_Photography from Pixabay.

Sources:

Jenkins DJ, Wolever TM, Taylor RH, et al. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr. 1981;34(3):362-366. doi:10.1093/ajcn/34.3.362

ScienceDirect. Glycemic Index. In Medicine and Dentistry. Accessed June 2026. https://www.sciencedirect.com/topics/medicine-and-dentistry/glycemic-index

ScienceDirect. Glycemic Load. In Medicine and Dentistry. Accessed June 2026. https://www.sciencedirect.com/topics/medicine-and-dentistry/glycemic-load

The University of Sydney. GI Search. Accessed June 2026. https://glycemicindex.com/gi-search/

Vega-López S, Venn BJ, Slavin JL. Relevance of the Glycemic Index and Glycemic Load for Body Weight, Diabetes, and Cardiovascular Disease. Nutrients. 2018;10(10):1361. doi:10.3390/nu10101361