The next diet in our series (following posts on the ketogenic diet and the Mediterranean diet) is the low glycemic index diet. While the first two diets are fairly well-defined and well-known, what constitutes a low glycemic index diet is less clear and the concept is much less well known.

But this post will clear things up!

First, I’ll explain what the glycemic index and the glycemic load are. Then I’ll talk about several studies that vary in how they implement a low glycemic index diet. From there, we’ll be able to see what features of the diet are most effective at helping with PCOS symptoms.

what is The glycemic index?

Blood sugar before, during, and after eating breakfast, as measured by a Signos continuous glucose monitor.

Every time you consume a food that contains carbohydrates, the concentration of sugar in your blood (what we refer to as your blood sugar) goes up until your body has a chance to either burn the sugar or store it. Once all of the carbohydrate you’ve consumed has been used or stored, your blood sugar returns to its baseline level. You can see the plot of blood sugar following breakfast in the photo here.

To calculate the glycemic index, researchers have people consume a reference food, usually white bread or straight glucose. They calculate the area under the blood sugar spike for this reference food, and this is then considered to be a glycemic index of 100. They then feed participants other foods containing the same grams of carbohydrates as the reference foods. For example, if the reference food was 100 g of glucose (commonly used), and the researchers were testing watermelon, they’d feed the test subjects about 9 cups of diced watermelon (there are 11 g of carbohydrates per 1 cup of diced watermelon).

While watermelon has a high glycemic index (76), it has a low glycemic load (8).

These other foods create smaller blood sugar spikes because they are digested more slowly. The area under their curves are then compared to the area under the curve created by the reference food. If the new area is 80% of the original area, the glycemic index for the tested food is 80 (Dodd et al. 2011). For watermelon, the area under the curve is 76% of the area for straight glucose, so it has a glycemic index of 76. For white rice, it is 72.

Relative to pure glucose, a glycemic index of 70 or higher is considered to have a high glycemic index, 56 to 69 is considered a medium glycemic index, and 55 or lower is considered to be low glycemic.

What is glycemic load?

The glycemic load recognizes that we generally eat different quantities of foods. When was the last time you ate 9 cups of diced watermelon? Maybe never? In contrast, there are about 50 g of carbohydrates in rice (varying a little depending on the type of rice). Have you ever eaten 2 cups of rice in one sitting? Maybe that last time you ordered a large plate of fried rice at your favorite Asian restaurant or a bowl of risotto at your favorite Italian restaurant?

White rice has both a high glycemic index (72) and a high glycemic load (35).

The glycemic load adjusts the glycemic index based on the amount of carbohydrates in a serving. The glycemic load for watermelon is only 8 while the glycemic load for white rice is 35 (Higdon et al. 2023).

While the glycemic index is probably better known, we really should pay more attention to glycemic load. This number better represents what happens in our bodies under realistic conditions. A glycemic load per serving of less than 10 is considered low, 10 to 20 is medium, and more than 20 is high.

As we’ll see below, many studies calculate total glycemic loads for their participants. This is simply summing up all the glycemic loads per serving of carbohydrates across all the servings they consumed. So if they consumed two servings of a food with a glycemic load of 10, 1 serving of a food with a glycemic load of 15, and 1 serving of a food with a glycemic load of 20, their total glycemic load for the day would be (2 x 10) + (1 x 15) + (1 x 20) = 55.

Why do we care about glycemic index and Glycemic load for PCOS?

The glycemic index and glycemic load indicate how hard our bodies have to work to bring blood sugar levels back to baseline after consuming a meal.

Women with PCOS tend to have insulin resistance (see an earlier post all about this!), which means that our bodies are not as good at bringing blood sugar levels back to baseline without producing a lot of insulin in the process. Our ovaries are also more sensitive to insulin and produce more testosterone in the presence of high insulin.

The more time our blood sugar remains elevated, the more insulin we secrete, and the more testosterone we secrete. Which means… more acne, more hair where we don’t want it, less hair where we do want it, and a ripple effect on our reproductive hormones that throw off our menstrual cycles.

Eating foods that have less of an impact on our blood sugar can help prevent this cascade of effects. So a low glycemic index (or low glycemic load!) diet makes a lot of sense for women with PCOS.

So now let’s explore what the literature says about its impacts!

THe literature on low glycemic index diets and PCOS

Unlike the previous two diets I’ve discussed in previous posts, there is a lot of high quality literature on the use of low glycemic diets for PCOS. As mentioned above, however, they vary in how they define low glycemic. I’ll pull out common themes at the end of this post.

A Low Glycemic index diet & insulin resistance, cardiovascular health, and reproductive hormones

This first study (Gower et al. 2013) demonstrates far-reaching effects of reducing the average glycemic index and load of the foods you eat. This study recruited 30 women with PCOS between the ages of 21 and 50 years old. Participants had BMIs less than 45, were not diabetics, and were not taking cholesterol medication, blood pressure medication, or oral contraceptives.

Half of the participants started the study with 8 weeks of a standard diet and the other half started with 8 weeks of a low carbohydrate diet. After the 8-week intervention period, all metrics were monitored. The participants then had 4 weeks without any diet intervention. Metrics were measured again, and then the groups switched diets, so the group that started with the standard diet now followed the low carbohydrate diet and vice versa. The second intervention period also lasted 8 weeks.

The standard diet included:

55% of calories from carbohydrates
18% of calories from protein
27% of calories from fat
An average glycemic index of 60 per carbohydate item consumed
A total glycemic load across all carbohydrates consumed ranging from 143 to 192 depending on total calorie consumption

The low carbohydrate diet included:

41% of calories from carbohydrates
19% of calories from protein
40% of calories from fat
An average glycemic index of 50
A total glycemic load across all carbohydrates consumed of 81 to 114
- By reducing both the glycemic index of foods and the proportion of calories coming from carbohydrates, this intervention led to a substantial reduction in glycemic load in the low carbohydrate diet.

Both groups consumed 1800 to 2500 calories per day, and the amount was determined based on individual energy needs.

Unlike other studies, this study monitored the effects of a liquid meal (Carnation Instant Breakfast mixed with whole milk) on blood sugar and insulin levels for four hours after consumption of the meal. By taking blood draws at regular intervals after the meal, they were able to determine how the different diets affect insulin secretion at different points after the meal.

The table below compares the results of the two diets. The standard diet had no effects on how the participants’ bodies responded to consuming a liquid meal relative to the response before following this diet. The low carbohydrate diet, however, did change how the body responded. The initial insulin response to consuming a meal was smaller, while the insulin response as blood sugar continued to rise from consuming the meal increased. Total insulin release remained unchanged.

The researchers looked at correlations between the different insulin response measures and found a positive correlation between the initial insulin release and testosterone levels, while the other measures of insulin release were not correlated with testosterone. This implies that the low carbohydrate diet’s effect on how much insulin is secreted immediately after consuming carbohydrates should help lower testosterone levels in women with PCOS and lessen PCOS symptoms.

And indeed, the low carbohydrate diet was associated with an almost 25% reduction in testosterone levels.

In addition to effects on post-meal insulin secretion and testosterone, the low carbohydrate diet reduced insulin resistance and improved insulin sensitivity and reduced cholesterol levels.

It should be noted that the low carbohydrate group replaced carbohydrate calories with fat calories. While fat consumption has been blamed for elevating cholesterol and triglyceride levels, we see the opposite effect with this study.

Healthy fats are healthy for us! Eat them!

A comparison of the effects of consuming a standard diet vs. a low carbohydrate diet for 8 weeks (Gower et al. 2013).

Low Glycemic index diet & frequency of Ovulation

While most studies focus on weight loss, reproductive hormones, and/or markers of insulin resistance, one study (Sordia-Hernández et al. 2016) considered the impact of a low glycemic index diet on frequency of ovulation.

This study recruited 40 women with PCOS. Half were assigned to a low glycemic index diet and the other half a high glycemic index diet. 19 of the women assigned to the low glycemic index diet participated for the entire duration of the study, while 18 from the low glycemic index diet participated for the entire duration.

Both groups consumed:

45 - 50% of their calories from carbohydrates
30 - 40% of their calories from fat
15 to 20% of their calories from protein
20 - 35 grams of fiber per day
between 1200 and 1500 calories per day

The low glycemic group consumed carbohydrates with glycemic indexes less than 45 while the high glycemic group consumed carbohydrates with glycemic indexes of 50 to 75.

The intervention lasted for 3 months. In each month, ovulation was monitored by ultrasound.

In the 1st month:

1 of 19 women in the low glycemic index diet had a confirmed ovulatory cycle
0 of 18 women in the high glycemic index diet had a confirmed ovulatory cycle

In the 2nd month:

6 of 19 women in the low glycemic index diet had a confirmed ovulatory cycle
1 of 18 women in the high glycemic index diet had a confirmed ovulatory cycle

In the 3rd month:

7 of 19 women in the low glycemic index diet had a confirmed ovulatory cycle

3 of 18 women in the high glycemic index diet had a confirmed ovulatory cycle

There are few takeaways from these results. First, the low glycemic index diet was associated with increased ovulation rates. This is especially pertinent for those trying to conceive.

Second, it look some time for the diet to affect ovulation. So if you try this diet, give it at least 2 months before you say it was ineffective. I like to give changes at least 3 months, as long as there are no negative side effects occurring.

Third, the diet did not work for everyone within the study time frame. It is unclear if more women in the low glycemic index diet group would have resumed ovulating if the study lasted for, say, 6 months instead of 3. I suspect that might be the case, but without data, we can’t say for sure.

This also highlights the heterogeneity of PCOS. What works for one person with PCOS may not work for another. I tend to think of PCOS tools as different levers. For some women, some levers are more effective than others. For other women, they can tradeoff between different levers more easily. I have major tradeoffs between stress and nutrition. If my stress levels are well-controlled, I can be more lax with nutrition. Under high stress, I can’t get away with as much nutrition-wise. Other women are need to focus more heavily on nutrition, regardless of stress levels, and vice versa. The women in this study who did not have success with nutrition might need to focus more on things like stress reduction or getting adequate sleep.

An emphasis on Glycemic Load instead of glycemic index

One study (Panico et al. 2014) considered the effects of reducing glycemic load while holding glycemic index constant. This study recruited 30 women with PCOS in Naples, Italy. It should be noted that only 7 completed the entire study. While this is low, their intervention spanned 6 months, and retaining participants for longer-term interventions is generally hard.

In this study, the participants served as their own controls. Half began with a low glycemic load diet while the other half began with a moderate to high glycemic load diet. Each group followed their respective diets for 3 months and then the groups switched the diets they were following.

The low glycemic load group consumed:

foods with an average glycemic index of 63 to 69
foods with total glycemic load amounting to 79 to 105 (varying with total calories consumed)
18% of their calories from protein
44% of their calories from carbohydrates
25 to 35 g of fiber per day

The moderately high glycemic load group consumed:

foods with an average glycemic index of 66 to 68

foods with total glycemic load amounting to 125 to 134 (varying with total calories consumed)
19 - 20% of their calories from protein
50 - 52% of their calories from carbohydrates
34 to 44 g of fiber per day (increased as a byproduct of increased carbohydrate consumption)

Both groups consumed 1500 to 1800 calories per day.

The table below summarizes the results found from this study. The study was designed to maintain weight, so it is unsurprising that weight loss and changes in BMI did not occur.

While fasting measures of insulin resistance did not indicate any substantial changes, measures taken two hours after eating demonstrated that the low glycemic load diet was improving how well the participants’ bodies handled carbohydrates after eating. There were no improvements seen after following the high glycemic load diet.

The low glycemic load diet also led to reductions in the two androgen hormones measured: testosterone and DHEA-S, which would likely lead to reductions in PCOS symptoms.

Lastly, this is the only study that looked specifically at stress hormones, and they found a decrease in adrenocorticotropic (ACTH) hormone, a hormone that triggers the release of stress hormones like cortisol, adrenaline, and noradrenaline. This reduction is notable because some women with PCOS have adrenal glands that are more sensitive to stress hormones like ACTH and produce excessive androgen hormones like testosterone in the presence of these stress hormones (Maas et al. 2016). Using diet to reduce stress hormones could help to less androgen excess in these women and improve PCOS symptoms.

A comparison of results after following either a low glycemic load diet or a moderately high glycemic load diet for three months (Panico et al. 2014).

Can a low glycemic diet lead to similar weight loss in women with PCOS and women without PCOS?

It is commonly understood that women with PCOS often have a harder time losing weight than women without PCOS. I strongly suspect that this is because foods that are commonly considered to be healthy (low-fat but sugar-sweetened yogurt topped with sugar-sweetened granola and berries, a big bowl of quinoa with vegetables, chickpeas, croutons, and possibly a little bit of chicken, or a plate of pasta with lean meat) are actually not so great for women with insulin resistance. You can cut calories and not lose weight if you body is always flooded with insulin, a hormone that tells your body to store fat.

So I was pretty excited to find a study (Shishegar et al. 2019) that compares weight loss outcomes across women with and without PCOS when both groups followed a low glycemic index diet.

This study recruited 28 women with PCOS and 34 without PCOS to follow a 24-week calorie restricted low glycemic index diet.

This diet included:

a 500 calorie deficit per day
50% of calories from carbohydrates
- carbohydrates restricted to low and medium glycemic index
20% of calories from protein
30% of calories from fat
lean meat, whole grains, low-fat dairy, non-starchy vegetables, vegetable oils

Participants received food menus based on their own energy requirements and eating habits.

The average glycemic index was about 41 and the average total glycemic load was about 66. Note that this study has a lower glycemic index and load than other studies discussed here, which likely contributes to the success of the intervention.

The table below shows the results for both groups of women. There are no significant differences in the percent changes in body weight, BMI, waist circumference, fasting insulin levels, or HOMA-IR (a measure of insulin resistance). The only metric for which the two groups differed was diastolic blood pressure. For this metric, the women with PCOS saw improvement in blood pressure while those without PCOS actually saw a slight increase in blood pressure.

A comparison of results for women with PCOS and women without PCOS after both groups followed a reduced calorie low glycemic index diet for 24 weeks (Shishegar et al. 2019).

I think these results should be really encouraging for women with PCOS who are trying to lose weight.

You can still eat 50% of your calories from carbohydrates and lose just as much weight as women without PCOS, as long as you choose your carbohydrate sources wisely!

how effective are small reductions in glycemic index and load?

One of the original studies (Marsh et al. 2010) to consider the effect of glycemic index and glycemic load on women with PCOS kept carbohydrate consumption the same (50% of calories) across their diet groups, but simply had participants in the low glycemic index group swap out higher glycemic index foods for lower glycemic index foods.

This study recruited 96 women with PCOS between the ages of 18 and 40. Participants were split into two groups:

A low-fat, low-GI diet group who consumed:

low-GI breads and cereals

A low-fat, conventional healthy diet group who consumed:

moderate to high GI breads and cereals

Both groups consumed:

moderate to high fiber foods
about 50% of their calories from carbohydrates
about 23% of their calories from protein
unsaturated fats like nuts, seeds, avocados, olive oil
- limited overall fat intake (27% of their calories)
high amounts of vegetables and salads
little added sugar or alcohol

Participants in both groups were instructed to follow their respective diets until they had lost 7% of their body weight or until 12 months had passed.

The low-GI diet group had an average glycemic index of 50 and total glycemic load of 82 while the conventional diet group had an average glycemic index of 59 and an average total glycemic load of 91. It should be noted that these actual values differ less between groups than the study had planned; compliance challenges resulted in the two groups eating more similarly than instructed.

Given the small difference in consumption between the two groups, it is not surprising that there were few statistically significant differences observed between the two groups.

The low-GI groups had a bigger reduction in insulin levels 2 hours after consuming a glucose drink, an indicator of reduced insulin resistance. However, none of other measures of insulin resistance or insulin sensitivity had statistically significant differences between the two diet groups.

There were no observed differences between weight loss, changes in hormones levels, or changes in measures of cardiovascular health.

Take-home points

Reducing the average glycemic index and glycemic load of the food you eat can reduce insulin resistance and increase insulin sensitivity as well as improve cardiovascular health.

Using a low glycemic index and/or load diet can result in just as much weight loss for women with PCOS as for women without PCOS.

Diets that were effective had average glycemic indexes ranging from 41 to 50. Having an average glycemic index of 63 to 69 was also effective when total carbohydrate consumption was reduced to 44% of calories.

Small changes in glycemic index and load may not be effective.

Diets that were effective at mitigating PCOS symptoms and health risks generally contained at least 30% of calories from fat.

The PCOS Carb Pyramid

This visual represents how I prioritize carb sources for both myself and for my PCOS clients. We can work in all carbs, but some carb sources should be consumed more frequently, and others only on occasion.

When thinking about my own diet and when working with clients, I always prioritize those foods that have both a low glycemic index and a low glycemic load. But there is very much a place for all foods, especially if they are foods you love! I like to think of carbs and the various combinations of glycemic load and glycemic index as a pyramid, much like the food pyramid that guided nutrition for decades. We want to focus using foods lower on the pyramid for our primary sources of carbohydrates, and eat those foods higher up less frequently. And we can still have an occasion treat from the top of the pyramid.

References

Dodd H, Williams S, Brown R, and Venn B. 2011. Calculating meal glycemic index by using measured and published food values compared with directly measured meal glycemic index. The American Journal of Clinical Nutrition, 94(4):992-996 https://doi.org/10.3945/ajcn.111.012138

Gower BA, Chandler-Laney PC, Ovalle F, Goree LL, Azziz R, Desmond RA, Granger WM, Goss AM, and Bates GW. 2013. Favourable metabolic effects of a eucaloric lower-carbohydrate diet in women with PCOS. Clinical Endocrinology 79:550–557 https://doi.org/10.1111/cen.12175

Higdon J. 2023. Glycemic Index and Glycemic Load. Linus Pauling Institute. Available: https://lpi.oregonstate.edu/mic/food-beverages/glycemic-index-glycemic-load#table-1

Maas KH, Chuan S, Harrison E, Cook-Andersen H, Duleba AJ, and Chang RJ. 2016 Androgen responses to adrenocorticotropic hormone infusion among individual women with polycystic ovary syndrome. Fertility and Sterility, 106(5):1252-1257 http://dx.doi.org/10.1016/j.fertnstert.2016.06.039

Marsh KA, Steinbeck KS, Atkinson FS, Petocz P, and Brand-Miller JC. 2010. Effect of a low glycemic index compared with a conventional healthy diet on polycystic ovary syndrome. American Journal of Clinical Nutrition 92:83–92 https://doi.org/10.3945/ajcn.2010.29261

Panico A, Lupoli GA, Cioffi I, Zacchia G, Caldara A, Lupoli G, Contaldo F, and Pasanisi F. 2014. Effects of an isocaloric low-glycemic-load diet in polycystic ovary syndrome. Nutritional Therapy & Metabolism 32(2):85-92 https://doi.org/10.5301/NTM.2014.12407

Shishehgar F, Mirmiran P, Rahmati M, Tohidi M and Tehrani FR. 2019. Does a restricted energy low glycemic index diet have a different effect on overweight women with or without polycystic ovary syndrome? BMC Endocrine Disorders 19:93 https://doi.org/10.1186/s12902-019-0420-1

Sordia-Hernández LH, Ancer Rodríguez P, Saldivar Rodriguez D, Trejo Guzmán S, Servín Zenteno ES, Guerrero González G, and Ibarra Patiño R. 2016. Effect of a low glycemic diet in patients with polycystic ovary syndrome and anovulation - a randomized controlled trial. Clinical and Experimental Obstetrics and Gynecology 43(4):555-559 https://doi.org/10.12891/ceog3037.2016