The cortisol correction for albumin is a calculation used to estimate the free (biologically active) cortisol fraction when total cortisol is measured in patients with low albumin levels, such as those with critical illness, nephrotic syndrome, or severe liver disease. Cortisol circulates predominantly bound to two carrier proteins: corticosteroid-binding globulin (CBG, also called transcortin), which binds approximately 80-90% of cortisol with high affinity, and albumin, which binds approximately 5-10% with lower affinity. Only the remaining 5-10% is free and biologically active. In critical illness, CBG is dramatically reduced (acute-phase response causes hepatic CBG synthesis to fall by up to 50%) and albumin also falls, leading to a paradoxical situation: total cortisol may appear low or 'normal' while free cortisol is actually substantially elevated due to reduced protein binding. Conversely, in chronic hypoalbuminaemia from nephrotic syndrome, total cortisol appears lower than it truly is in terms of adrenal reserve, because less is protein-bound. The Hamrahian formula estimates free cortisol as: Free cortisol = Total cortisol x (1 - 0.9 x albumin/(albumin + Kd)), where Kd (the dissociation constant for albumin-cortisol binding) is approximately 3.6 g/dL in most validated versions. This correction is important because both under-diagnosis of adrenal insufficiency (missing a low free cortisol when total is 'normal') and over-diagnosis (treating patients for AI when total cortisol is low only because of hypoalbuminaemia) can cause significant patient harm.
Free Cortisol (estimated) = Total Cortisol x (1 - 0.9 x Albumin / (Albumin + 3.6)); where Albumin in g/dL and Kd = 3.6 g/dL
- 1Measure total serum cortisol and serum albumin from the same blood sample.
- 2Express albumin in g/dL (if in g/L, divide by 10).
- 3Calculate the albumin correction factor: 0.9 x Albumin / (Albumin + 3.6).
- 4Subtract the correction factor from 1.0 to get the free fraction.
- 5Multiply total cortisol by the free fraction to obtain estimated free cortisol.
- 6Interpret: In critical illness, if total cortisol appears borderline but free cortisol is adequate (above 2 mcg/dL free), adrenal insufficiency is less likely. If free cortisol is below 1.8 mcg/dL, adrenal insufficiency may be present despite a 'normal' total cortisol.
- 7Recognise that direct free cortisol measurement by equilibrium dialysis is the gold standard but is rarely available in routine clinical laboratories.
Low albumin means more cortisol is free — total underestimates adrenal function
This patient's total cortisol of 350 nmol/L might seem insufficient (borderline by some critical care criteria). However, with severe hypoalbuminaemia (18 g/L), 70% of cortisol is free, yielding an estimated free cortisol of 245 nmol/L — well above the threshold for adequate adrenal response. This patient likely does not have adrenal insufficiency.
Despite low total cortisol, free cortisol suggests normal HPA axis response
Nephrotic syndrome causes urinary loss of albumin and CBG. Total cortisol appears low, raising suspicion of adrenal insufficiency, but the free cortisol estimate of 190 nmol/L is consistent with normal HPA function. The Short Synacthen Test with both total and free cortisol measurement would confirm.
Normal albumin — total cortisol adequately reflects adrenal function
When albumin is normal, approximately 47% of cortisol is estimated as free. The correction changes the interpretation modestly but does not dramatically alter clinical conclusions. Total cortisol is a reliable indicator of adrenal function in patients with normal albumin.
Very high free cortisol — appropriate stress response, no steroid supplement needed
A total cortisol of 900 nmol/L in septic shock with low albumin yields an estimated free cortisol of 593 nmol/L — a robust stress cortisol response. Supplemental hydrocortisone is unlikely to add benefit from an adrenal reserve perspective, though it may be indicated for haemodynamic reasons in refractory septic shock.
Assessing adrenal function in critically ill ICU patients with hypoalbuminaemia to avoid inappropriate steroid supplementation. This application is commonly used by professionals who need precise quantitative analysis to support decision-making, budgeting, and strategic planning in their respective fields
Interpreting cortisol results in patients with nephrotic syndrome to prevent missed diagnosis of genuine adrenal insufficiency. Industry practitioners rely on this calculation to benchmark performance, compare alternatives, and ensure compliance with established standards and regulatory requirements
Correcting misleading total cortisol values in patients with liver cirrhosis before adrenal testing decisions. Academic researchers and students use this computation to validate theoretical models, complete coursework assignments, and develop deeper understanding of the underlying mathematical principles
Supporting endocrinology consultations when total cortisol results appear inconsistent with the clinical picture. Financial analysts and planners incorporate this calculation into their workflow to produce accurate forecasts, evaluate risk scenarios, and present data-driven recommendations to stakeholders
Research studies investigating the true prevalence of relative adrenal insufficiency in critical illness. This application is commonly used by professionals who need precise quantitative analysis to support decision-making, budgeting, and strategic planning in their respective fields
Septic shock and critical illness
Septic shock causes the most dramatic falls in CBG and albumin. Total cortisol thresholds used for diagnosing relative adrenal insufficiency in septic shock (e.g., peak cortisol below 550 nmol/L post-ACTH, or increment below 250 nmol/L) were derived from studies that did not universally measure albumin or free cortisol. Albumin correction is strongly recommended before diagnosing adrenal insufficiency in septic shock.
Liver failure
Severe hepatic cirrhosis reduces albumin synthesis and CBG synthesis, both of which are produced in the liver. Total cortisol falls while free cortisol may be preserved or even elevated, particularly with adrenal hyperfunction that accompanies portal hypertension and high circulating ACTH. Total cortisol in liver failure is particularly unreliable without albumin correction.
Cushing syndrome assessment
In Cushing syndrome investigation, very high cortisol production may mask true binding protein status. Total cortisol is usually markedly elevated, and the free fraction — while proportionally higher — is also substantially elevated. Late-night salivary cortisol (which measures free cortisol in saliva) and urinary free cortisol are preferred over serum free cortisol estimation in Cushing syndrome workup.
Exogenous glucocorticoid use
Patients on exogenous glucocorticoids may have suppressed endogenous cortisol production with variable total cortisol depending on recent dosing. Prednisolone cross-reacts with cortisol assays (hydrocortisone most, dexamethasone least). Timing of the blood sample relative to the last steroid dose is critical to avoid assay interference.
| Albumin (g/L) | Albumin (g/dL) | Free Fraction (%) | Estimated Free Cortisol (nmol/L) |
|---|---|---|---|
| 40 (normal) | 4.0 | 52.6% | 263 |
| 35 | 3.5 | 54.0% | 270 |
| 25 | 2.5 | 62.5% | 313 |
| 20 | 2.0 | 67.9% | 340 |
| 15 | 1.5 | 75.0% | 375 |
| 10 (very low) | 1.0 | 84.4% | 422 |
Why does critically ill patients' total cortisol underestimate free cortisol?
In critical illness, both CBG and albumin fall dramatically. CBG is an acute-phase negative reactant — its hepatic synthesis decreases during the stress response. With less carrier protein available, a greater proportion of circulating cortisol is unbound (free) and biologically active. Total cortisol therefore underestimates the physiologically available free fraction.
Is there a validated formula for free cortisol estimation?
The Hamrahian formula (2004) is the most widely cited: Free cortisol = Total cortisol x (1 - 0.9 x albumin/(albumin + Kd)), where Kd = 3.6 g/dL. The Coolens formula uses a more complex calculation incorporating CBG concentration when it is measured. Both are estimates; direct free cortisol measurement by equilibrium dialysis remains the gold standard.
What is the normal range for free cortisol?
Normal morning free serum cortisol by equilibrium dialysis is approximately 0.7-2.0 mcg/dL (20-55 nmol/L), corresponding to about 7-10% of total morning cortisol. In critical illness, a free cortisol above 1.8 mcg/dL (50 nmol/L) is generally considered indicative of an adequate adrenal stress response. In practice, this concept is central to cortisol correction because it determines the core relationship between the input variables.
Should every ICU patient have cortisol corrected for albumin?
Cortisol-albumin correction is particularly important when critically ill patients are being assessed for possible relative adrenal insufficiency (RAI) and have low albumin (below 25 g/L). In patients with normal albumin, total cortisol is adequate. The ACTH stimulation test (Short Synacthen Test) is still the most widely used clinical assessment, and albumin correction adds interpretive nuance.
Can urinary free cortisol be used instead of serum free cortisol?
Urinary free cortisol (UFC) reflects the fraction of cortisol filtered by the kidneys and excreted unbound in urine, which correlates with serum free cortisol over a 24-hour period. However, UFC is used for diagnosing Cushing syndrome rather than for acute critical illness assessment, where a rapid result from serum is needed.
What effect does CBG on cortisol measurement?
CBG (corticosteroid-binding globulin / transcortin) binds cortisol with high affinity. If CBG is measured directly, the Coolens equation can calculate free cortisol more precisely. In clinical practice, CBG is rarely measured routinely. The albumin-only correction is a simplified approximation that accounts for albumin's role but cannot fully account for variable CBG levels.
Does ethnicity affect the Kd constant?
The dissociation constant (Kd = 3.6 g/dL) used in the Hamrahian formula was derived from predominantly Western populations. Limited data suggest that the cortisol-albumin binding constant may vary slightly with ethnicity and with specific albumin genetic variants, but these differences are unlikely to be clinically significant in routine practice. This is an important consideration when working with cortisol correction calculations in practical applications.
What are the limitations of the albumin correction formula?
The formula assumes constant CBG levels, which is incorrect in critical illness (CBG falls markedly). It uses a fixed Kd derived from healthy volunteer studies. It does not account for post-translational CBG modifications that alter its binding affinity in critical illness. Free cortisol estimated from this formula should be considered an approximation; direct measurement by equilibrium dialysis is always preferred when clinically feasible.
Pro Tip
When albumin is below 25 g/L, multiply the apparent total cortisol by approximately 1.3-1.5 as a quick mental estimate of corrected free cortisol. This rough correction helps avoid mistakenly diagnosing adrenal insufficiency in hypoalbuminaemic patients before formal calculation.
Did you know?
Corticosteroid-binding globulin (CBG/transcortin) is so efficient at binding cortisol that it acts as a natural glucocorticoid reservoir. At sites of inflammation, locally produced elastase cleaves CBG, rapidly releasing bound cortisol exactly where the anti-inflammatory effect is needed most — a remarkable example of targeted drug delivery by the body itself.
References
- ›Hamrahian AH et al. Measurements of serum free cortisol in critically ill patients. NEJM 2004
- ›Annane D et al. Assessment of adrenal function in the critically ill: plasma cortisol vs free plasma cortisol. Crit Care 2006
- ›Nguyen HB et al. Critical illness and adrenal function. J Intensive Care 2016
- ›Bornstein SR et al. Diagnosis and Treatment of Primary Adrenal Insufficiency (Endocrine Society 2016)