Types of PCOS: What the Research Actually Shows About How It Presents
If you’ve ever wondered about the different types of PCOS, you’re not alone. PCOS is a complex condition that presents quite differently from person to person.
A major clustering study published in late 2025 (Nature Medicine) identified four distinct subtypes of PCOS based on clinical and biochemical features. This builds on earlier large-scale genome-wide association studies from 2020 that also pointed to underlying heterogeneity in the condition. Importantly, most people with PCOS show a combination of features rather than fitting neatly into just one subtype.
Because our understanding of PCOS subtypes has advanced significantly in recent years, much of the following information may be new to you.
By Dr. Fiona McCulloch, ND Author of 8 Steps to Reverse Your PCOS and peer reviewer for the 2023 International Evidence-Based Guidelines for the Assessment and Management of Polycystic Ovary Syndrome.
You read about PCOS on social media before, recognized yourself, followed the protocol, and nothing changed. The instinct is to assume that you are just harder to treat. Neither of those assumptions is usually correct. We have learned a great deal about the endocrinology of how PCOS actually behaves in individuals. For treatment to be successful the individual endocrinology of PCOS must be understood.
A wave of large-scale genomic research is now showing why treatments for PCOS often don’t work: the underlying factors in PCOS are often not properly identified, can overlap, co-occur, and shift over time. If the standard framework for PCOS left you without a clear answer or results, this is why.
What Genome-Wide Research Reveals About PCOS Presentations
Two landmark studies form the basis for the pattern framework on this page. The patterns overlap, and it is common to have a combination of features rather than one clear presentation. That is the finding, not a caveat. Daye et al. (2020, PLOS Medicine, PMID 32574161) used supervised clustering analysis to identify distinct phenotypic subgroups with novel genetic associations. Gao et al. (2025, Nature Medicine, PMID 41162652) expanded this work with nearly 12,000 women across five international cohorts. Four recognizable patterns emerged. This is preliminary research, not finalized diagnostic categories, and Dr. Fiona McCulloch frames it that way explicitly, because that is what the data supports.
Before reading the subtypes below, one thing matters: these are tendencies and current states, not permanent identities. The goal of reading what follows is not to find your type and stop there. It is to identify which factors appear to be present in your picture. Most people will recognize themselves in more than one subtype. That is by design, and it is what the research actually shows. Read all four before drawing any conclusions.
Dr. Fiona McCulloch, ND has been assessing and treating PCOS as overlapping individual factors for over 24 years. When the clustering studies were published, they confirmed what clinical observation had already shown, not the other way around. The 2023 International PCOS Guidelines, which she peer-reviewed, reflect the same move toward individualized, factor-based assessment of a person’s unique endocrinology.
The Insulin-Resistant Pattern
Of the four patterns, the insulin-resistant presentation has the most clearly mapped treatment pathway. When insulin resistance is correctly identified and treated directly, research suggests this pattern has the highest remission potential of all four. That qualifier matters: correctly identified. Not assumed, not inferred from body weight, and not ruled out by a single fasting blood test. The recognition and the caveat belong together, and separating them produces the same incomplete picture that brought many readers to this page.
Common features of this pattern include weight gain concentrated around the midsection, difficulty losing weight despite dietary changes, skin tags, darkened patches of skin at the neck or underarms (acanthosis nigricans), elevated fasting insulin, and a metabolic profile that trends toward higher triglycerides and lower HDL cholesterol. The associated genetic signal involves chromosome 2, in the KCNH7-FIGN-GRB14 region. Over time, this pattern carries elevated risk for type 2 diabetes and fatty liver disease if insulin resistance is not addressed. That risk is real, and it is also modifiable. Patients can treat insulin resistance and shift substantially out of this dominant presentation. It is not a fixed destination.
Here is where the picture gets complicated for many readers: standard fasting insulin and fasting glucose tests can come back completely normal while post-meal insulin is significantly elevated. Based on clinical observation at White Lotus Clinic, approximately 50% of lean PCOS patients whose fasting tests showed normal results have shown elevated post-meal insulin on more sensitive testing using the Kraft protocol. (That figure is a clinical estimate, not a published statistic.) If a previous test said your insulin was normal, it may mean the standard test was not sensitive enough to catch what is happening after a meal, not that insulin resistance is ruled out. The reader who just recognized herself in this subtype should finish it thinking: this fits me, but was my testing actually adequate?
Insulin also directly signals the ovaries to produce testosterone. Elevated insulin lowers sex hormone-binding globulin (SHBG), which raises the amount of testosterone circulating freely in the body. This is the mechanism connecting the insulin-resistant pattern to elevated androgens, and it is why most people with the IR pattern also have at least moderately elevated androgens. If you recognized yourself here and also have androgenic symptoms such as acne, hair thinning, or excess body hair, both patterns may be active simultaneously. The combination of IR and high androgens is the most common co-occurring pair in PCOS clinical practice.
The High SHBG Pattern (Lean PCOS)
If you followed insulin-focused advice and nothing changed, this is usually why. Approximately 30% of people with PCOS have no clinically significant insulin resistance. If you are lean, if your metabolic numbers look unremarkable, and if you have been cutting carbohydrates or restricting food because every PCOS resource assumed insulin was the issue, this is the pattern that explains why that did not work. Without insulin resistance driving the presentation, the metabolic risks associated with this pattern are lower, and management tends to be more straightforward, though the underlying hormonal dynamics are still worth understanding in detail. The genetic signals here involve BMPR1B and PRDM2 genes, which relate to ovarian follicle signaling. Elevated SHBG (sex hormone-binding globulin) is a liver protein that binds to testosterone and keeps it biologically inactive. When SHBG is high, more testosterone is bound, less is free to act on tissues. This pattern typically includes elevated LH, higher SHBG, and a relatively lean body composition. If you have PCOS and have been told that losing more weight or cutting carbohydrates further should help, but it has not, this pattern may explain why: the primary driver is not insulin.
DHEA-S is a weak androgen produced by the adrenal glands. Elevated DHEA-S is sometimes interpreted as adrenal PCOS, a label that has circulated widely. But the interpretation depends heavily on what the rest of the androgen picture looks like. If DHEA-S is elevated while testosterone is lower-normal, SHBG is high and biological androgen activity is not especially high, this may actually place a patient closer to the high SHBG pattern than a distinctly androgenic one. Whether DHEA-S is converting to stronger androgens matters far more than its raw level. Hidden insulin resistance is also possible even in lean patients: normal fasting results do not rule it out. Two layers of uncertainty exist here simultaneously, and both point toward the same conclusion: the full picture requires a full assessment.
The High SHBG pattern and the High LH-AMH pattern frequently overlap. Elevated LH is common in both, and it is the connecting feature. LH drives follicle activity in the ovary and influences both SHBG modulation and AMH production through shared ovarian signaling pathways. If this subtype fits you, the next one probably also has something to tell you. Do not stop here.
The High LH / High AMH Pattern
If you have been told your AMH is very high, or that you have polycystic ovaries with a large number of follicles, there is some good news. Based on research and clinical observation: the High LH-AMH pattern is actually associated with a large follicle count, or ovarian reserve. That reserve is typically associated with a later natural menopause, and with the capacity to conceive later in life compared to the general population. For a reader who has been anxious about high AMH or polycystic ovaries, this can be a relief. The feature causing anxiety may also be a fertility advantage. The presentation includes strong ovarian activity, high LH, very high AMH, and a large follicular reserve. Because AMH is secreted by ovarian follicles, very high levels can slow follicle maturation and reduce sensitivity to FSH, which is why ovulation may be delayed or irregular. A genetic link has been identified for this subtype at the BMPR1B gene, which encodes an AMH receptor. AMH can be measured in a standard blood test without an ultrasound, which matters for readers still in the early stages of investigation.
This pattern does not tell the full androgen story. Some patients with the high LH-AMH presentation have completely normal androgen levels, making it a very specific pattern where the primary feature is ovarian reserve rather than androgen excess. Others have elevated androgens through co-occurrence with the androgenic pattern. AMH alone does not reveal the androgen picture. If you have recognized yourself in this subtype, you do not yet know whether androgens are also a factor. That is the question this subtype leaves open, intentionally.
Two named overlaps are worth noting. First, the High LH-AMH pattern and the High SHBG pattern frequently co-occur. Elevated LH connects them. If the previous subtype resonated with you at all, these two patterns may both be relevant. Second, high AMH can accompany the high androgenic pattern. AMH tends to remain high even with treatment in many cases, improving to a more “high-normal level”. Patients with this type will always have a higher than normal ovarian reserve though the level of AMH improves with treatment.
The High Androgenic Pattern (Adrenal DHEA-S and Ovary Testosterone)
This pattern includes high levels of androgenic hormones: testosterone and sometimes DHEA-S (adrenal). DHEA-S is a weak androgen, therefore a patient is thought o have the androgenic pattern if they have high testosterone, or stronger androgens like DHT. Many also have high DHEA-S, : this is more common in younger women. It’s also possible to have the androgenic pattern with high testosterone and normal DHEA-S.
Of the four patterns, high androgens come with some genuine biological advantages that rarely appear in standard PCOS content. People with the high androgenic pattern tend to gain muscle mass more easily than those with other presentations. There are also noted benefits for cognition and drive. Later menopause and ability to conceive a bit later is typical. The pattern responds well to targeted treatment.
The visible symptoms, including acne, hirsutism, and hair loss of the androgenic type, are real and often can be extremely stressful. Androgenetic alopecia, the pattern of hair thinning associated with androgen excess at the scalp, is gradually progressive and can cause a lot of distress for patients. Strategies can slow progression and support existing hair. Hirsutism or unwanted hair growth is particularly common in this type. Other symptoms include cystic acne, oily skin and dandruff.
No specific genetic markers have been identified for this pattern in genome-wide studies, possibly because androgen excess is so common across all PCOS presentations that it does not generate a distinctive genetic signature of its own. That absence is itself informative, and it points toward the most important thing about this subtype.
Most PCOS patients across all four patterns have at least moderately elevated androgens. This is not a feature unique to the high androgenic pattern. It cuts across presentations the way inflammation does, near-universally, as a feature of the condition rather than a differentiator between categories. The genome-wide research likely did not find a distinctive genetic signal for this pattern precisely because androgen elevation is too common to separate one cluster from the others. What varies is the degree, the source, and the specific combination with other features. Most subtypes also carry androgen-related features. The degree and combination are what vary. The absence of androgen involvement is actually the less common finding in PCOS, though the most common type you’ll see this in is the LH-AMH type..
Two mechanistic links are worth naming. First, insulin resistance drives androgen elevation through a direct pathway: high insulin signals the ovaries to produce more testosterone, while simultaneously lowering SHBG. This is a common combination in clinical practice, IR and high androgens co-occurring, connecting the first subtype directly to the last. Second, DHEA-S, which is produced by the adrenal glands, sometimes behaves differently in perimenopause. In younger women with PCOS, elevated DHEA-S fits within the androgenic pattern or the high SHBG pattern depending on the full picture. In perimenopause, DHEA-S tends to rise relative to other hormones through a different mechanism, producing more testosterone as a downstream effect. A reader in her late 30s or 40s experiencing new androgenic symptoms should not assume the same PCOS pattern she had at 25 is still the primary driver. The picture shifts with life stage.
So at this point, you may recognize parts of yourself in several subtypes, and still not know what to do next. That is not a failure of attention. It is the point.
Why Most People With PCOS Have a Combination of Patterns
The problem is not that you have not figured out your type. It is that type is not the right level of thinking.
The genome-wide research says this directly. “It is common to have a combination of features rather than one clear type of PCOS.” That is not a hedge or a marketing softening. It is the finding. Some features run through all four subtypes in ways that prevent any single section from completing the picture. First, moderately elevated androgens for age appear in most PCOS patients regardless of which pattern is dominant, though in certain patients this may be more notable. Inflammation, is a common feature of PCOS, and can be found in many subtypes. See also: inflammation in PCOS.
Patterns also shift over time. A patient who treated her insulin resistance effectively may no longer have that as the dominant driver. What remains might be elevated AMH and some residual androgenic features. The earlier framework may have been accurate at the time. The picture is different now. Dr. Fiona McCulloch’s own clinical thinking on “adrenal PCOS” has evolved with clinical experience: the same treatment approaches that work for androgen excess generally work for adrenal androgen excess too, which argues against it being a truly separate category. When the person who wrote the book on PCOS patterns updates her own position, that is not contrarianism. It is what the data showed.
If you recognized yourself across more than one pattern above, or could not settle cleanly on any single one, that is the most common presentation, not an unusual one. The framework was not built to produce one clean answer. It was built to show why clean answers are the problem.
A note for conversations with other practitioners: recent clustering studies, including a 2025 Nature Medicine study of nearly 12,000 women across five international cohorts, show that PCOS presents as overlapping patterns with combinations of features, not as four mutually exclusive categories. This is preliminary data, but it is large-scale, peer-reviewed, and published in one of the leading medical journals. If you need language to explain why a type-based approach may not have worked for you, that framing is now grounded in published research rather than a practitioner’s personal framework.
The question worth asking now is not which type applies to you. It is which factors are active in your presentation, in what specific combination, right now. That question has a different kind of answer, one that requires a full picture of what is actually happening, not a self-identification with the closest available category.
If reading about subtypes of PCOS raised more questions, that is what assessment is designed to clarify. Here is what identifying your specific combination actually looks like.
Questions This Page Actually Answers
I have already tried supplements and dietary changes based on my PCOS type. Nothing worked. Why would this approach be any different?
The most common reason PCOS treatments do not work is that they are based on a generic approach. PCOS requires specific, targeted treatments rather than protocols applied to a broad category. Sometimes the pattern being treated is not quite right. Sometimes the right pattern is being treated but the interventions are not potent enough: gentle supplementation may not be sufficient when androgens are genuinely very high or insulin resistance is clinically significant. PCOS is an intense condition and sometimes requires more than a gentle approach. The failure is usually structural, not personal. The starting point that changes outcomes is a full picture of which factors are actually driving symptoms, not which label comes closest. “I tried everything” often means “I tried everything that matched the type I assumed I had.”
Is the overlapping patterns framework actually supported by research, or is this just another practitioner's own system?
The framework described on this page comes directly from two peer-reviewed studies: Daye et al. (2020, PLOS Medicine, PMID 32574161) and Gao et al. (2025, Nature Medicine, PMID 41162652). The Gao study involved nearly 12,000 women across five international cohorts using genome-wide association methods. Dr. Fiona McCulloch frames these findings as preliminary data, not finalized diagnostic categories, because that is what they are. Where other practitioners overstate the certainty of clustering findings, the willingness to hedge builds more credibility here, not less. When the claims about a framework are appropriately qualified, the specific findings within that framework become more trustworthy.
Can you really have more than one PCOS pattern at the same time?
Yes. Combination of features is actually the most common presentation in PCOS, not the exception. The most frequent co-occurring pair in clinical practice is the insulin-resistant and high androgenic patterns, connected through the direct pathway where elevated insulin raises testosterone production and lowers SHBG. The High LH-AMH and High SHBG patterns also overlap frequently, with elevated LH as the connecting mechanism. Most PCOS patients across all four patterns have at least moderately elevated androgens. Single-pattern presentations exist but they are not the majority. Assuming your PCOS is one thing is usually where the self-management gap starts.
My fasting insulin was normal. Does that rule out insulin resistance?
Not on its own. Standard fasting insulin and fasting glucose tests measure insulin levels at rest, after an overnight fast. Insulin resistance in PCOS often manifests most clearly after eating: insulin rises higher than normal and stays elevated longer in response to food, but the fasting level looks unremarkable. Based on clinical observation, approximately 50% of lean PCOS patients whose fasting tests were normal have shown elevated post-meal insulin on more sensitive testing using the Kraft protocol. (That figure is a clinical observation estimate, not a published statistic.) A normal fasting test result does not mean insulin resistance is not present. It may mean the standard test was not sensitive enough to find it. More on how insulin resistance is tested in PCOS.
I am lean and do not have obvious metabolic symptoms. Does the pattern framework still apply to me?
Yes, and the High SHBG pattern was built precisely to describe presentations like yours. Approximately 30% of people with PCOS have no clinically significant insulin resistance. The High SHBG pattern describes lean patients specifically, with its own genetic signals (BMPR1B, PRDM2) and its own hormonal dynamics. Lean patients are as genuinely affected by PCOS as anyone else, through different mechanisms: ovarian signaling, androgenic activity, AMH-driven follicle dynamics. The pattern framework applies more precisely to your situation than any model built primarily around insulin resistance and weight, because it was developed to describe exactly the kinds of presentations that do not fit the IR-dominant narrative.
How PCOS Factors Are Actually Identified
- FSH
- LH
- Total testosterone
- Free testosterone
- SHBG
- DHEAS
- Fasting insulin
- Fasting glucose
- HbA1C
- Lipid panel
- Kraft insulin resistance test
- Liver enzymes
- AMH
The reason prior approaches did not work was usually not personal. A generic protocol cannot target what it has not identified. The supplement that is helpful for insulin-resistant PCOS may be irrelevant for a lean patient with high AMH and normal insulin, and irrelevant treatments at best produce nothing and at worst distract from what would actually help. Once the specific factors driving symptoms are identified, treatment can be matched to what is actually happening. That is the structural difference between a general approach and a specific one. For PCOS, the specific one is the one that works.
Meet our medical director, Dr. Fiona McCulloch, ND
Dr. Fiona McCulloch, ND is the author of 8 Steps to Reverse Your PCOS (HarperCollins), a clinical guide for patients and practitioners that has shaped how PCOS is approached in naturopathic and integrative care. She served as a peer reviewer for the 2023 International PCOS Guidelines and has maintained a clinical focus in PCOS assessment and treatment for over 24 years. The clustering research described on this page confirms patterns she has been observing and treating clinically for decades.
- Author: 8 Steps to Reverse Your PCOS (HarperCollins)
- Peer reviewer: 2023 International PCOS Guidelines
- Studies cited on this page: Daye 2020 (PLOS Medicine), Gao 2025 (Nature Medicine)
- Clinical PCOS focus: assessment and treatment, 24+ years
The question is not which type you are. It is which factors are active in your presentation, in what combination, right now. That question has a specific answer, and identifying it is where treatment actually starts.
Dr. Fiona McCulloch, ND and the White Lotus Clinic team see patients in Ontario.
References
- Daye M, Jones GL, Ledger WL, et al. Distinct Subtypes of Polycystic Ovary Syndrome with Novel Genetic Associations: A Supervised Phenotypic Clustering Analysis. PLOS Medicine. 2020. PMID: 32574161.
- Gao H, et al. Genome-wide association study identifies four biologically distinct subtypes of polycystic ovary syndrome. Nature Medicine. 2025. PMID: 41162652. (Five international cohorts, approximately 12,000 participants.)