Sachi Health Can a low-starch, low-dairy diet lower insulin and testosterone in PCOS? What a study of 24 women found
In a study of 24 women with PCOS, eight weeks of a low-starch, low-dairy diet, with no calorie counting and no medication, was linked to about a 50% drop in fasting insulin and lower testosterone and hirsutism scores.
Twenty-four women with PCOS cut starches and most dairy from their meals for eight weeks: no grains, no beans, no added sugar, no milk or yogurt (a little cheese was allowed). They didn't count calories and weren't told to exercise more. By week eight, their fasting insulin had dropped by about half, they'd lost an average of 8.6 kg, and their testosterone and hirsutism scores had both come down. The whole point was to eat in a way that keeps insulin low, since high insulin is one of the engines driving PCOS symptoms.
The numbers that matter
Can we actually trust this?
Partly, and it's worth being clear about why. This was a small study: 24 women finished it, and there was no control group. Everyone followed the same diet, and everyone was compared only against their own "before" numbers. That design can show that things changed, but it can't fully prove what caused the change.
The bigger catch is that every participant lost weight, and weight loss on its own improves insulin and androgen levels in PCOS. So the diet, the weight loss, and the lower carb intake are all tangled together; the researchers say this directly, and they couldn't separate one effect from another. What this study does well is show a realistic, food-based approach that produced meaningful improvements in just eight weeks without medication. Treat it as a promising signal worth trying, not as settled proof that this specific diet beats every other one.
Does this apply to you?
This study is most relevant if:
- You have PCOS and were diagnosed by the Rotterdam criteria (the most widely used definition)
- You're carrying extra weight, since participants started with a BMI between 25 and 45
- You deal with insulin-related symptoms, excess hair growth (hirsutism), or both
- You'd rather adjust <em>what</em> you eat than count every calorie; this was an eat-until-satisfied approach
Why this matters
If you have PCOS, you've probably been told to "just lose weight" without anyone explaining how, or why it's harder for you than it seems to be for everyone else. That advice lands like a closed door.
What's quietly hopeful about this study is the order of operations: the women didn't starve themselves or train for a marathon. They changed the type of food on their plates, ate until they were full, and the numbers that actually drive PCOS, namely insulin, testosterone, and hair growth, moved in the right direction within two months.
The hirsutism improvement is small on paper, but for many women the visible symptoms are the ones that wear hardest on day-to-day life, and seeing them respond to food alone is the part worth holding onto.
If you want to try this
- Build meals around protein and non-starchy vegetables.
Lean meat, poultry, fish, eggs, and plenty of vegetables were the foundation of every plate in this study.
- Cut the starches and added sugars.
That means no grains, no beans, and no added sugar, including honey, agave, and fruit juice from concentrate.
- Pull back on dairy, but keep a little cheese if it helps.
Participants dropped milk and yogurt but were allowed up to 1 oz of full-fat cheese a day, since cheese spikes insulin less than other dairy and made the diet easier to stick to.
- Lean on low-sugar fruit, nuts, seeds, and healthy fats.
Berries, apples, oranges, avocado, olives, nuts, and olive or coconut oil were all on the menu.
- Don't count calories: eat until you're satisfied.
The participants weren't restricting portions. The lever here is food <em>type</em>, not hunger.
- Talk to your doctor first if you take medication.
Participants paused insulin sensitizers and hormonal birth control under medical supervision before starting; this isn't something to do on your own.
FAQ
- Can changing your diet really lower testosterone in PCOS?
- In this study it did. After eight weeks of cutting starches and most dairy, free testosterone dropped 23% and total testosterone dropped 19%, with no medication involved. The likely chain is that lowering insulin lowers the androgens that drive PCOS symptoms. Testosterone is the main androgen behind things like excess hair growth and acne, so bringing it down tends to ease those symptoms.
- Is cutting dairy good or bad for PCOS?
- This study cut most dairy because milk and yogurt raise insulin more than their sugar content alone would suggest, and high insulin is a central driver of PCOS symptoms. Cheese was the exception: participants were allowed up to 1 oz a day because it raises insulin less than other dairy. This was a low-dairy approach paired with low starch, so it can't tell you dairy alone is the culprit, but it's a reasonable lever to test.
- Will a low-starch diet help hirsutism (unwanted hair growth)?
- Possibly, slowly. Hirsutism is male-pattern hair growth (face, chest, back) driven by elevated androgens, measured here with the modified Ferriman-Gallwey score. That score improved by 2.1 points over eight weeks, which is real but small. Hair growth responds slowly because of how hair cycles work, so this is a longer game than the weight or insulin numbers, which moved faster.
- Do I need to have a certain type of PCOS for this to apply?
- The participants were diagnosed by the Rotterdam criteria, the most common PCOS definition, which requires at least two of: irregular periods, signs of high androgens, and polycystic ovaries on ultrasound. They were also overweight or obese (BMI 25 to 45) with insulin and androgen issues. If that sounds like you, the approach is most likely to apply. If your PCOS is lean or not insulin-driven, the insulin-lowering logic may matter less for you.
How Sachi can help
Sachi is built to run a protocol like this one for you, not just record it after the fact. The diet in this study comes down to a handful of repeating goals (keep starches low, skip added sugar, build meals around protein and vegetables), and those goals come pre-loaded in the app, so you're not typing out what you ate and did every single day. You set the goals once, and Sachi tracks your consistency against them.
Paired with nutrition logging, that means you can see whether you're actually following the pattern this study used, and, once symptom and measurement tracking is in your hands, whether your own numbers are responding the way theirs did. Some of this is live today and some is still in development, but the goal is simple: make a research-informed plan something you can follow and measure, instead of guess at.
Study at a glance
Reference
Phy JL, Pohlmeier AM, Cooper JA, Watkins P, Spallholz J, Harris KS, Berenson AB, Boylan M (2015). Low Starch/Low Dairy Diet Results in Successful Treatment of Obesity and Co-Morbidities Linked to Polycystic Ovary Syndrome (PCOS). Journal of Obesity & Weight Loss Therapy , 5(2), 259. https://doi.org/10.4172/2165-7904.1000259
Cite this issue
Sachi Health. (June 30, 2026). Can a low-starch, low-dairy diet lower insulin and testosterone in PCOS? What a study of 24 women found. The PCOS Brief. https://www.sachi-health.com/blog/the-pcos-brief-issue-17
Further reading from The PCOS Brief
Nasiri M, Monazzami A, Alavimilani S, Asemi Z (2024). Modulation of hormonal, metabolic, inflammatory and oxidative stress biomarkers in women with polycystic ovary syndrome following combined (resistant and endurance) training: a randomized controlled trial. BMC Endocrine Disorders, 25(1), 1. https://doi.org/10.1186/s12902-024-01793-0 — Read our plain-English summary.
Barrea L, Arnone A, Annunziata G, Muscogiuri G, Laudisio D, Salzano C, Pugliese G, Colao A, Savastano S (2019). Adherence to the Mediterranean Diet, Dietary Patterns and Body Composition in Women with Polycystic Ovary Syndrome (PCOS). Nutrients, 11(10), 2278. https://doi.org/10.3390/nu11102278 — Read our plain-English summary.