Low SHBG in PCOS: Why It Matters More Than Total Testosterone

What SHBG does and why low SHBG raises free testosterone

The liver makes SHBG specifically to bind circulating testosterone and oestrogen, holding roughly 95–98% of total testosterone in an inactive, protein-bound state at any given moment. Only the small unbound fraction — free testosterone — can act on tissue. When SHBG falls, less testosterone stays bound, so free testosterone rises even if total testosterone has not moved at all. A 2022 review of SHBG’s role in PCOS describes this as the mechanism behind a common and confusing lab pattern: total testosterone inside the reference range, but clear clinical signs of excess androgen.

Note: PCOS was renamed polyendocrine metabolic ovarian syndrome, or PMOS, in May 2026, by a global consensus of more than 50 organisations. The SHBG mechanism behind androgen symptoms is unchanged — only the name on the diagnosis is. This article uses PCOS, since that is still the term most readers search.

The free androgen index: why it matters more than total testosterone

The free androgen index (FAI) converts total testosterone and SHBG into a single number that tracks the biologically active hormone more closely than either value alone: FAI = (total testosterone ÷ SHBG) × 100. The calculation was formalised in a 1999 evaluation of methods for estimating free testosterone, which found SHBG-based calculations tracked directly-measured free testosterone far more reliably than total testosterone did on its own. The 2023 international PCOS guideline endorses calculated free testosterone or the FAI over total testosterone alone when assessing biochemical hyperandrogenism for exactly this reason.

The number only means something next to SHBG. Two women can share an identical total testosterone of 55 ng/dL and land in very different places once SHBG is factored in.

The identical total testosterone reads two different ways depending on SHBG — which is the whole argument for testing SHBG alongside testosterone rather than testosterone alone.

Insulin is the reason SHBG falls

Insulin acts directly on the liver to suppress SHBG production, and PCOS is disproportionately a condition of elevated circulating insulin, which is why low SHBG shows up so often in the insulin-resistant phenotypes. In a study of 854 women with PCOS, an SHBG below 42 nmol/L predicted a meaningfully higher probability of both insulin resistance and impaired fasting glucose than higher SHBG values did — making a single SHBG number a rough, low-cost signal for a metabolic risk that would otherwise need a separate glucose-insulin work-up to see.

Reading your number by phenotype

SHBG’s behaviour tracks phenotype closely. In the insulin-resistant phenotypes, SHBG is typically at its lowest, free testosterone and the FAI at their highest, and glucose handling most often abnormal on testing — the pattern the 854-woman study above describes. In the lean phenotype, SHBG is often normal or even elevated, since there is less circulating insulin to suppress it; total and free testosterone can sit closer together, and androgen elevation is more often driven by LH output than by the SHBG mechanism. That difference is a real, checkable reason two people with the same total testosterone can have very different visible symptoms — more free testosterone reaching hair follicles and skin in one case than the other.

Genetics set part of your baseline, independent of insulin

Not all of your SHBG number comes from your current metabolic state. The 2022 review of SHBG’s role in PCOS also describes variation in the SHBG gene itself contributing independently to how high or low a given person’s baseline sits, separate from body weight or insulin resistance — one reason two women with similar insulin levels and similar weight can still land on meaningfully different SHBG results. This does not change what to do about a low number, since the actionable lever is still the insulin side of the equation, but it is worth knowing before assuming a low result reflects lifestyle or weight alone. It also explains why some women with lean, insulin-sensitive PCOS still run a lower SHBG than expected for their metabolic profile, and why SHBG alone is a supporting data point rather than a verdict on either phenotype or effort.

What a low SHBG result does not tell you

SHBG is not specific to PCOS, and a low value on its own does not diagnose anything. Obesity alone lowers SHBG independent of PCOS, thyroid disease and some liver conditions shift it in both directions, and — in the other direction — hormonal contraception containing oestrogen raises SHBG substantially, which can mask a genuinely low baseline in anyone tested while on the pill. That last point matters for interpretation: an SHBG drawn on combined hormonal contraception tells you very little about the SHBG you would have off it, so a truly low baseline may only become visible after stopping, under your clinician’s guidance.

Who this doesn’t work for

A low SHBG will not confirm PCOS in someone whose weight alone explains it — obesity suppresses SHBG through the same insulin pathway regardless of the ovaries, so the number cannot distinguish “PCOS with obesity” from “obesity alone” without the rest of the diagnostic panel, and the full diagnostic criteria, alongside it. It is also uninterpretable in anyone currently on oestrogen-containing hormonal contraception, and it will not explain androgen symptoms in the lean, LH-driven phenotype, where SHBG is often unremarkable despite real biochemical hyperandrogenism.