💜
PCOS and weight loss: what works and what is being sold
Research review · PubMed
✍️ Автор: Катя
«Your metabolism is broken», «cut the carbs», «inositol will fix your cycle», «dairy and gluten make it worse». We took the 2023 international guideline — a document built by thousands of specialists under a rigorous methodology — and checked every popular claim against it. A great deal of what is sold as mandatory is either not recommended there, or actively contradicted.
🧊
Myth one: «it is harder to lose weight with PCOS»
The headline finding: **it is not**. When women with and without PCOS were put on the same programme, they lost weight **equally** (−3.9 kg vs −4.5 kg). Basal metabolic rate is also **the same** — 1520 vs 1464 kcal. The guideline states it plainly: there is «a lack of consistent evidence of physiological differences». **Your metabolism is not broken.**
📈
But something genuinely is harder
Following 14,000 women over 19 years showed that women with PCOS gained **4.6 kg more**. The honest framing is this: **losing weight is not harder — keeping it off is**. And the same dietary slip-ups cost more.
🎯
Myth two: «you need to lose 5–10%»
That target **does not appear in the 2023 guideline at all**. It was replaced by an individualised approach. What emerged instead is more interesting: the effect is **dose-dependent with no threshold**. Every **1%** of weight lost added **5.6%** to the odds of ovulating. Without weight loss 19% of women ovulated; with any loss at all, more than half did. It starts working from the first kilograms — there is no magic ten percent to wait for.
🍽️
Myth three: «there is a special PCOS diet»
The guideline's exact wording: **«there is no evidence to support any 1 type of diet composition over another»** — not for weight, not for hormones, not for cycles. And a separate clause requires clinicians to **avoid unduly restrictive and nutritionally unbalanced diets**.
🥑
Keto: the numbers fall apart on inspection
This is the most striking part. Loud meta-analyses promise 9–10 kg. But they **pool proper trials together with studies that had no comparison group at all**. When only the **controlled** data are analysed, the picture changes: weight loss is **0.52 kg and not statistically significant**, and testosterone shows no effect either. The only thing that genuinely improves is insulin resistance. Those «minus 10 kg» are an ordinary harsh calorie deficit, not carbohydrate magic.
🌾
What does give a small genuine benefit
The best candidate is a **low glycaemic index** approach. It lowered testosterone with almost no variability between studies — the cleanest hormonal signal in the entire PCOS diet literature. And in a one-year trial where **both groups lost the same amount of weight**, menstrual regularity returned in **95% versus 63%**. So carbohydrate quality does matter — just far more modestly than advertised.
🏃♀️
Movement is one of the few «strong» recommendations
Concrete numbers from the guideline: **150–300 minutes** of moderate activity a week to avoid gaining weight, and **250 minutes** if the goal is losing it. Plus **strength work twice a week**. Choose whichever type you enjoy — no form has been shown superior. Higher-intensity training returns more metabolically per minute, and benefits appear **even without weight change** — that is a separate clause in the guideline.
😴
The unexpected part: sleep outranks every dietary recommendation
Obstructive sleep apnoea affects roughly **35%** of women with PCOS — and crucially, the **certainty of that evidence is higher** than for any recommendation about diet or supplements. The guideline explicitly requires screening: **snoring + waking unrefreshed + daytime sleepiness**. Untreated apnoea directly worsens insulin sensitivity, which undermines everything you do with food.
💊
Inositol: it improves your labs, not the thing you took it for
The most popular supplement. The honest picture: it genuinely **does improve** insulin resistance and fasting insulin. But ovulation — **no significant effect**; pregnancy — **no significant effect**; weight and hirsutism — likewise. The guideline lists it as «**could be considered**» at low to very low certainty, and states that **metformin should be preferred**. Inositol's real advantage is tolerability: roughly ten times fewer gastrointestinal side effects.
⚖️
And that fashionable 40:1 ratio
It is supported by **exactly one** study at «very low» certainty. The guideline itself writes that «specific types, doses, or **combinations** of inositol **cannot currently be recommended**». There is no basis for paying extra for a number on the label.
💉
Metformin is not a weight-loss drug
The precise figure: against placebo it changes body mass index by **0.48** — for a woman of 165 cm, roughly **1.3 kg**. The guideline requires telling women that «**metformin and active lifestyle intervention have similar efficacy**». Its indication is metabolism and cycles, not weight. For weight there is a **separate** recommendation — anti-obesity medication, with roughly **four to five times** the effect, and always **alongside** lifestyle change. The document separately warns of the **high risk of weight regain after discontinuation**.
🧠
And now the strongest evidence in the whole document
It is not about diet and not about supplements. The **only recommendations at the highest certainty** in the entire guideline are **screening for depression and anxiety** in every woman with PCOS. For comparison, every recommendation about metformin and exercise sits in the **lowest** tier. The document separately requires considering **eating disorders regardless of weight** — especially during weight management. In other words, the science says mental health in PCOS is better established than any diet.
🚩
Also: «adrenal PCOS» does not exist
Splitting PCOS into «insulin / inflammatory / adrenal / post-pill types» is an **internet classification, not a medical one**. Across 254 recommendations the words «cortisol» and «adrenal type» appear **zero times**, and there is no recommendation to test cortisol. Elevated adrenal androgens do occur in roughly 20–30% of women — but that is **a finding on a lab report, not a separate diagnosis with separate treatment**.
🚫
Five food myths — none held up
**«Dairy makes PCOS worse»** — the data point the other way: full-fat dairy is associated with **less** anovulation, skimmed with more. **Gluten** — a total of three papers exist in the literature, none of them a proper trial. **«Soy disrupts hormones»** — the opposite: in a trial it **lowered** insulin and free androgens. **«Coffee is harmful»** — an analysis of 1.1 million people shows diabetes risk **falling** with each cup. **Detox teas** — zero studies, and senna, their usual base, is linked to serious bowel complications.
🗣️
And about language — this is in the guideline too
There is a **dedicated section on weight stigma**. The terms «overweight» and «obesity» are recognised as stigmatising; «**higher weight**» is recommended instead. Clinicians are advised to ask permission before weighing and to offer a **choice** between a weight-centric and a weight-inclusive approach. And the strongest line: weight may turn out to be a **non-modifiable** factor when lifestyle is the only available tool.
✅
What to actually do
**1.** A moderate calorie deficit by whatever method you can sustain for years — composition is secondary. **2.** Movement: 150–300 min, or 250 if the goal is weight loss, plus strength twice weekly. **3.** Get sleep apnoea checked if you snore and wake unrefreshed. **4.** Medication with a doctor: metformin for metabolism, combined pills for hormones, separate drugs for weight. **5.** Inositol optional, not mandatory. **6.** Do not spend money on detoxes, gluten-free protocols or cutting dairy, soy and coffee. **7.** If dieting brings binges and guilt, that is a signal to stop and ask for help — not to «tighten the discipline».
❓ Часті запитання
Is it really harder to lose weight with PCOS?
On the same programme, women with and without PCOS lose weight equally, and basal metabolic rate does not differ. Keeping it off is genuinely harder: over 19 years the difference came to about 4.6 kg.
How much do I need to lose for my cycle to return?
There is no specific target in the 2023 guideline. The effect is dose-dependent and starts with the first kilograms: each percent of weight lost adds roughly 5.6% to the odds of ovulating.
Do I need keto or a very low-carb diet?
No. The guideline states plainly that no diet composition has been shown superior. In controlled trials keto produced a statistically non-significant weight loss; the large figures come from studies with no comparison group.
Does inositol work?
It improves markers of insulin resistance, but no significant effect on ovulation, pregnancy, weight or hirsutism has been shown. The guideline lists it as optional and prefers metformin.
Should I cut dairy and gluten with PCOS?
There is no evidence for it. For dairy the data actually run counter to expectation, and trials of gluten-free diets in PCOS essentially do not exist. The guideline advises against unduly restrictive diets.
📚 Дослідження, згадані в статті (14)
- KEY: 2023 international evidence-based guideline for PCOS (JCEM / Monash / ESHRE / ASRM) — джерело →
- Inositol in PCOS: review for the 2023 guideline — 30 RCTs, 2,230 women (JCEM, 2024) — джерело →
- Lifestyle change in PCOS: low-quality evidence (Cochrane, 2019) — джерело →
- Keto in PCOS: controlled vs uncontrolled data (Reprod Biol Endocrinol, 2025) — джерело →
- Diets in PCOS under isocaloric designs: no advantage (Nutrients, 2025) — джерело →
- Low glycaemic index: testosterone and metabolism, 10 RCTs (Advances in Nutrition, 2021) — джерело →
- One-year RCT: low GI vs conventional diet — cycles 95% vs 63% (Am J Clin Nutr, 2010) — джерело →
- Exercise in PCOS: intensity matters more than modality (Frontiers in Physiology, 2020) — джерело →
- Sleep apnoea and metabolic abnormalities in PCOS (Sleep, 2018) — джерело →
- Metformin in PCOS: guideline meta-analysis — BMI −0.48 (Clinical Endocrinology, 2019) — джерело →
- Metformin vs lifestyle: similar efficacy (Human Reproduction Update, 2015) — джерело →
- Dairy and anovulatory infertility: 18,555 women (Human Reproduction, 2007) — джерело →
- Soy isoflavones in PCOS: RCT showed improvement (JCEM, 2016) — джерело →
- Coffee and type 2 diabetes risk: 1.1 million participants (Diabetes Care, 2014) — джерело →
Це не медична порада. Маєш діагноз чи обмеження — звернися до лікаря.
Чи була стаття корисною?