The Free Androgen Index (FAI) is a calculated biochemical marker used in clinical medicine to estimate the proportion of unbound, biologically active testosterone in the bloodstream, particularly in assessing androgen status.[1] It is derived from the formula FAI = (total testosterone in nmol/L ÷ sex hormone-binding globulin [SHBG] in nmol/L) × 100, providing an indirect measure of free testosterone without directly assaying it.[2] This index accounts for the fact that approximately 98% of circulating testosterone is bound to proteins like SHBG and albumin, rendering it inactive, while the unbound fraction exerts physiological effects such as promoting muscle growth, bone density, and secondary sexual characteristics.[3]FAI is primarily employed as a diagnostic tool for hyperandrogenism in women, where elevated levels may indicate conditions like polycystic ovary syndrome (PCOS), hirsutism, or androgen-secreting tumors, often presenting with symptoms such as excess hair growth, acne, or menstrual irregularities.[4] In men, it can help evaluate hypogonadism or low testosterone-related issues like fatigue and reduced libido, though its utility is more limited due to higher baseline androgen levels and greater SHBG variability.[5] Reference ranges vary by age, sex, and laboratory method, but typical values for adult women are 0.4–8.4 (or 0.18–7.07% in some assays), while for men they range from 14.0–128.0, with measurements ideally taken in the morning to account for diurnal fluctuations.[2]Despite its widespread use, FAI has limitations, as it may overestimate free testosterone in cases of low SHBG (e.g., due to obesity or insulin resistance) and is less accurate than direct free testosterone assays via equilibrium dialysis, particularly in complex scenarios like pregnancy or liver disease.[1] It is typically measured via immunoassays on serum samples and serves as a cost-effective initial screening tool in endocrinology, guiding further tests like imaging or genetic analysis when abnormalities are detected.[4]
Overview
Definition
The free androgen index (FAI) is a calculated ratio used in endocrinology to estimate the levels of bioavailable testosterone in the blood, providing an indirect measure of the unbound, physiologically active form of this hormone.[6] Androgens are a class of steroidhormones responsible for the development and maintenance of male characteristics, with testosterone serving as the primary androgen in both men and women.[7] In circulation, testosterone exists in two main forms: bound and free. The bound fraction, which constitutes the majority (approximately 98%), is attached to carrier proteins such as sex hormone-binding globulin (SHBG) with high affinity (about 45-60%) and albumin with lower affinity (about 35-50%), rendering it biologically inactive and unavailable to target tissues.[8] In contrast, the free fraction, comprising only 1-2% of total testosterone, remains unbound and is the active form capable of diffusing into cells to exert its effects on gene expression and physiological processes.The FAI approximates this free testosterone fraction by accounting for variations in SHBG levels, which can significantly influence androgen availability without altering total testosterone concentrations.[9] This index is particularly valuable in clinical settings where direct measurement of free testosterone—via methods like equilibrium dialysis—is often impractical due to technical challenges, low concentrations, and assay inaccuracies in routine laboratories. By offering a simpler surrogate, the FAI facilitates the assessment of androgen status in conditions involving altered protein binding, such as those affecting liver function or hormone regulation.[10]
Historical Development
The free androgen index (FAI) was originally introduced in the early 1980s as a surrogate for direct measurement of free testosterone, specifically to assess androgen status in women with hirsutism, where early assays for unbound testosterone were technically demanding, costly, and often unreliable due to interference and low sensitivity.[11] This index, calculated as the ratio of total testosterone to sex hormone-binding globulin (SHBG), provided a practical approximation of bioavailable androgens without requiring complex equilibrium dialysis or ultrafiltration methods prevalent at the time.[11]Key early publications validated FAI's clinical utility in hyperandrogenism evaluation. For instance, a 1987 study by Wilke and Utley examined FAI alongside total and calculated free testosterone in hirsute women, finding it elevated in 41-68% of cases, outperforming total testosterone alone in detecting subtle androgen excess.[12] Subsequent research in the late 1980s and early 1990s further established FAI as a reliable marker for conditions involving altered SHBG binding, such as in polycystic ovary syndrome (PCOS), amid growing recognition of its correlation with clinical symptoms like hirsutism.By the 2000s, FAI had become a standard tool in endocrinology, endorsed in influential guidelines for hyperandrogenism assessment. The 2003 Rotterdam consensus criteria for PCOS incorporated biochemical hyperandrogenism evaluation, where FAI is commonly used as a metric to assess elevated androgens alongside clinical signs.[13] Later international evidence-based guidelines, such as those from 2018 and updated in 2023, explicitly recommended FAI as a simple and validated estimate of free testosterone for PCOS diagnosis, reflecting its widespread adoption in clinical practice.[14]
Calculation and Interpretation
Formula
The free androgen index (FAI) is calculated using the formula:\text{FAI} = \left( \frac{\text{total testosterone}}{\text{SHBG}} \right) \times 100where total testosterone and sex hormone-binding globulin (SHBG) are both measured in nmol/L, resulting in a unitless index that is often interpreted as a percentage-like value.[15] This simple ratio provides an indirect estimate of the biologically active fraction of testosterone in circulation.[16]The derivation of the FAI stems from the physiological binding dynamics of testosterone in blood plasma. Testosterone circulates primarily bound to proteins: approximately 40-60% to SHBG with high affinity, 40-50% to albumin with lower affinity, and only 1-3% remains unbound as free testosterone, which is the active form available to tissues.[15] Since SHBG binding dominates and varies significantly under physiological conditions (e.g., influenced by estrogen levels or liver function), the free fraction is inversely related to SHBG concentration. To approximate free testosterone, the formula divides total testosterone by SHBG concentration, normalizing for this binding effect; the multiplication by 100 scales the result for practical interpretation, yielding values typically ranging from 0.5-10 in adult women and 20-120 in adult men.[16] This approach assumes a linear relationship where changes in SHBG primarily dictate the free fraction, providing a step-wise simplification: first, measure total testosterone as the sum of bound and free forms; second, divide by SHBG to estimate the unbound proportion; third, scale by 100 to express as an index.[17]A key assumption in the FAI formula is the neglect of albumin binding, which is treated as relatively constant and of lower impact due to albumin's weaker affinity constant (approximately 3.6 × 10^4 L/mol) compared to SHBG (1 × 10^9 L/mol).[15] This simplification holds reasonably well in scenarios where SHBG fluctuations are the main variable, such as in women with hyperandrogenism, but may introduce inaccuracies when albumin levels vary (e.g., in liver disease).[16]Interpretation of the FAI focuses on deviations from expected values: an elevated FAI suggests increased free testosterone availability, indicating potential hyperandrogenism, while a low FAI points to reduced free testosterone, suggestive of hypoandrogenism.[17] This index thus serves as a practical surrogate for assessing androgen status without direct measurement of free testosterone, which requires more complex methods like equilibrium dialysis.[15]
Reference Ranges
The free androgen index (FAI) is a unitless ratio expressed as a percentage, calculated using total testosterone and sex hormone-binding globulin (SHBG) measured in consistent units, typically nmol/L in the SI system, to ensure accurate results across laboratories; using conventional units like ng/dL for testosterone and mg/dL for SHBG requires appropriate conversion factors to maintain equivalence, though SI units are standard for this computation.[18][19]Reference ranges for FAI are established by major clinical laboratories based on healthy populations and vary slightly by assay method and demographics, with endocrine societies recommending consultation of local laboratory values for interpretation.[20] For adult women, typical ranges are 0.7-8.7% in premenopausal individuals (ages 19-50 years) and 0.5-4.7% in postmenopausal women (age ≥50 years), reflecting lower androgen activity post-menopause.[21][4] In adult men, ranges are higher at 24.5-113.3% (ages 19-50 years) and 19.3-118.4% (age ≥50 years), corresponding to greater baseline androgen levels.[21] Optimal values from population data align closely, with 0.19-3.63% for women and 35-92.6% for men.[5]In children and adolescents, FAI ranges exhibit age- and sex-specific patterns due to pubertal development, remaining low in prepubertal stages and rising during puberty; comprehensive pediatric intervals from the CALIPER cohort confirm these dynamics, with values generally below 1% in young children and increasing to adult-like ranges by late adolescence.[22] Representative ranges from NHS laboratories illustrate this progression:
Age Group
Females (%)
Males (%)
<1 year
0.0-1.3
0.0-32.7
1 to <9 years
0.0-1.3
0.0-0.6
9 to <14 years
0.1-2.6
0.2-34.7
14 to <19 years
0.6-6.5
3.6-83.3
These ranges assume standardized conditions, as FAI exhibits some diurnal variation similar to total testosterone, with higher levels in the morning; samples are typically collected in the morning (8-11 a.m.) to align with peak levels and enhance reproducibility. Fluctuations with menstrual cycle phase are minimal.[4][21]
Laboratory Measurement
Testing Procedure
The testing procedure for the free androgen index (FAI) begins with the collection of a venous blood sample, typically 1 mL of serum obtained using a red-top or gel-barrier tube.[2]Serum is preferred over plasma for this measurement to ensure compatibility with standard assays for total testosterone and sex hormone-binding globulin (SHBG).[2] Samples should be collected in the morning, ideally between 7 a.m. and 10 a.m., to account for the diurnal variation in testosterone levels, which peak early in the day.[23]Fasting is not required prior to the blood draw.[24]Patient preparation is minimal but includes informing the healthcare provider of all medications, herbs, vitamins, and supplements being taken, as certain substances can temporarily alter hormone levels.[25] Specifically, patients on high-dose biotin therapy (greater than 5 mg/day) should discontinue it at least 72 hours before sample collection to avoid interference with immunoassay results.[2] Avoidance of acute conditions like severe illness or recent intense exercise is also recommended, as these can cause short-term fluctuations in measured values.[5]Once collected, the serum sample is processed to measure total testosterone and SHBG using laboratory assays. Total testosterone is commonly quantified via electrochemiluminescence immunoassay (ECLIA), a type of automated immunoassay, though liquid chromatography-tandem mass spectrometry (LC-MS/MS) is used in reference laboratories for greater specificity and accuracy, particularly in cases of low concentrations.[2][26] SHBG levels are similarly measured by immunoassay methods, such as ECLIA, which provide reliable quantification in serum.[2] These assays are performed on automated platforms to ensure precision and reproducibility.Following the individual measurements, the FAI is calculated automatically by laboratory software or manually if needed, using the values of total testosterone and SHBG obtained from the assays.[2] The resulting index value is then reported alongside reference ranges tailored to the patient's age and sex for contextual evaluation.[27]
Factors Influencing Measurement
Several endogenous factors can significantly influence the free androgen index (FAI) by altering levels of sex hormone-binding globulin (SHBG) or total testosterone, the components used in its calculation. Aging is associated with increased SHBG concentrations in both men and women, which inversely affects FAI by reducing the proportion of free androgens.[28]Obesity, particularly central adiposity, is linked to decreased SHBG levels due to insulin resistance and hyperinsulinemia, thereby elevating FAI.[29] Liver diseases such as cirrhosis often elevate SHBG production in hepatocytes, leading to lower FAI values.[30] Thyroid dysfunction also plays a role; hyperthyroidism increases SHBG, decreasing FAI, while hypothyroidism reduces SHBG, potentially raising FAI.[30]Exogenous factors, including medications and physiological states, further modulate FAI measurements. Oral contraceptives substantially increase SHBG through estrogenic effects, resulting in lowered FAI, often by 50% or more.[31] Glucocorticoids, such as dexamethasone, typically suppress SHBG levels, which can elevate FAI by increasing the unbound fraction of testosterone.[32] During pregnancy, SHBG levels rise markedly due to elevated estrogen, suppressing FAI throughout gestation.[33] In menopause, SHBG concentrations show no significant overall change compared to premenopause, though individual variations influenced by body composition may subtly affect FAI.[34]Assay-related variability introduces additional challenges in FAI measurement. Inter-laboratory differences in total testosterone and SHBG immunoassays can lead to substantial discrepancies in calculated FAI, with coefficients of variation up to 16% for low testosterone levels and 10% for SHBG.[35] These inconsistencies arise from variations in assay methodologies, calibration standards, and reagentsensitivity across labs.[36]Temporal physiological variations also impact FAI reliability. Circadian rhythms cause total testosterone to peak in the early morning and decline by up to 30-40% throughout the day, while SHBG remains relatively stable, thus affecting FAI diurnally.[37] Seasonal fluctuations in androgen levels are observed, with total testosterone and FAI often lower in spring (e.g., March) compared to autumn (e.g., August-October), potentially by 10-15%, influenced by photoperiod and environmental factors.[38]
Clinical Applications
Diagnosis of Hyperandrogenism in Women
Hyperandrogenism in women manifests through clinical symptoms such as hirsutism, acne, androgenic alopecia, and irregular menstrual cycles, where the free androgen index (FAI) serves as a key biochemical marker to confirm elevated androgen activity.[39] These symptoms arise from excess androgens stimulating hair follicles and sebaceous glands, often prompting evaluation for underlying endocrine disorders.[40] FAI, calculated from total testosterone and sex hormone-binding globulin levels, provides an estimate of bioavailable testosterone, aiding in distinguishing hyperandrogenism from other causes of these symptoms.[6]Diagnostic thresholds for FAI in women typically indicate hyperandrogenism when values exceed 5%, supporting a biochemical diagnosis alongside clinical findings.[6] For instance, an FAI greater than 4.5 has been used in some studies for identifying androgen excess, particularly in reproductive-age women presenting with hirsutism.[41] Similarly, thresholds around 5% or higher correlate with increased free testosterone fractions, helping to quantify the severity of androgen elevation.[6] These cutoffs are applied after confirming normal reference ranges for women, typically lower than in men due to physiological differences in androgen production.FAI is integrated into a multimodal diagnostic approach, combining biochemical results with clinical scoring systems like the modified Ferriman-Gallwey (mFG) score, which quantifies hirsutism by assessing hair growth in nine androgen-sensitive areas (score ≥8 suggesting excess).[42] An elevated mFG score prompts FAI testing to corroborate clinical hyperandrogenism, while pelvic imaging, such as transvaginal ultrasound, may be added if severe symptoms or markedly raised androgens suggest ovarian pathology.[42] This combined evaluation enhances diagnostic accuracy, avoiding reliance on any single metric.The Endocrine Society recommends calculating free testosterone from total testosterone and sex hormone-binding globulin (SHBG) levels as part of the initial evaluation for androgen excess in women with suspected hyperandrogenism, alongside total and free testosterone measurements in guidelines for hirsutism assessment.[42] Specifically, for women with moderate to severe hirsutism or additional signs like menstrual irregularities, early morning serum testing including such calculations helps identify treatable causes efficiently.[42] This approach aligns with broader recommendations for systematic screening to guide management and rule out rare tumors or adrenal disorders.[42]
Role in Polycystic Ovary Syndrome
The free androgen index (FAI) plays a key role in the diagnosis of polycystic ovary syndrome (PCOS) under the Rotterdam criteria, which require the presence of at least two of three features: oligo- or anovulation, clinical or biochemical hyperandrogenism, and polycystic ovarian morphology on ultrasound. Biochemical hyperandrogenism is assessed using FAI, calculated as the ratio of total testosterone to sex hormone-binding globulin (SHBG), to identify elevated androgen levels when clinical signs like hirsutism are absent or equivocal. This integration allows FAI to serve as objective evidence supporting the hyperandrogenism criterion, facilitating accurate PCOS classification in reproductive-age women. However, the 2023 International Evidence-based Guideline for PCOS prioritizes total and free testosterone assays for biochemical hyperandrogenism, recommending calculated indices like FAI only if reliable direct free testosterone measurement is unavailable.[43][39][44]Elevated FAI is prevalent in 70-80% of PCOS cases, reflecting underlying hyperandrogenemia and aiding differentiation from other hyperandrogenic conditions such as non-classic congenital adrenal hyperplasia or idiopathic hirsutism, where FAI levels may not align with polycystic ovarian features. Studies have demonstrated strong correlations between elevated FAI and polycystic ovarian morphology, with higher FAI values associated with increased antral follicle counts and ovarian volume, indicating a direct link to the syndrome's reproductive phenotype. Additionally, FAI positively correlates with insulin resistance in PCOS, as hyperinsulinemia suppresses SHBG production, further elevating free androgen bioavailability and exacerbating metabolic disturbances.[45][46][47][48]In PCOS management, FAI serves as a biomarker for monitoring treatment efficacy, with reductions observed following anti-androgen therapies such as spironolactone or combined oral contraceptives, which lower testosterone levels and restore hormonal balance. Lifestyle interventions, including weight loss through diet and exercise, also decrease FAI by improving insulin sensitivity and increasing SHBG, thereby alleviating hyperandrogenic symptoms and supporting long-term metabolic health. These changes in FAI levels provide quantifiable feedback on therapeutic response, guiding adjustments in PCOS care plans.[49][50][51]
Validity and Limitations
Accuracy as a Free Testosterone Proxy
The free androgen index (FAI) serves as an indirect estimate of free testosterone levels, primarily through its correlation with calculated free testosterone derived from total testosterone, sex hormone-binding globulin (SHBG), and assumed constant albumin concentrations. Studies have demonstrated strong correlations between FAI and calculated free testosterone, with Pearson correlation coefficients typically ranging from 0.8 to 0.9 in women with suspected hyperandrogenism.[6] However, FAI shows weaker agreement with the gold standard method of equilibrium dialysis for measuring free testosterone, where correlations are often lower (r ≈ 0.6-0.8) due to differences in accounting for binding dynamics.[52] In one validation study of 147 women, FAI correlated with equilibrium dialysis-measured free testosterone at r = 0.93, but this high value was specific to normal-range SHBG levels and did not hold across broader populations.[52]A key limitation of FAI arises from its underlying assumptions, particularly the neglect of variations in serum albumin concentration, which binds approximately 36% of testosterone with low affinity. In states of elevated albumin, such as certain inflammatory or nutritional conditions, FAI overestimates free testosterone because it fails to adjust for increased nonspecific binding, leading to inflated estimates of bioavailable hormone.[1] Similarly, FAI underestimates free testosterone in individuals with SHBG genetic variants that alter binding affinity, such as the Asp327Asn polymorphism, which reduces SHBG's testosterone-bindingcapacity and results in higher actual free fractions than predicted by concentration alone.[53] These assumptions contribute to inaccuracies at extremes of binding protein levels, with FAI overestimating calculated free testosterone by up to 20-30% when SHBG is low, a common scenario in hyperandrogenic states.[54]Meta-analyses have affirmed FAI's reliability as a proxy in women, particularly for distinguishing hyperandrogenism in polycystic ovary syndrome, with pooled sensitivity and specificity exceeding 80% across ethnic groups.[55] Nonetheless, variability in SHBG immunoassays introduces inconsistency, with inter-assay coefficients of variation up to 15%, amplifying errors in FAI calculations.[6] Despite these imperfections, FAI remains widely preferred over direct methods like equilibrium dialysis or ultrafiltration due to its low cost, rapid computation from routine total testosterone and SHBG assays, and sufficient clinical utility in resource-limited settings, though experts recommend confirmatory testing with mass spectrometry-based approaches for ambiguous cases.[52]
Applicability in Men and Special Populations
In adult men, the free androgen index (FAI) exhibits limited validity for evaluating androgen status, particularly in the diagnosis of hypogonadism, owing to greater variability in sex hormone-binding globulin (SHBG) levels compared to women. The FAI, calculated as (total testosterone / SHBG) × 100, relies on assumptions from the law of mass action that SHBG binding capacity far exceeds testosterone concentration, an assumption that does not hold in men where testosterone levels are substantially higher. A seminal 1993 study reported a weak correlation (r = 0.435) between FAI and directly measured free testosterone (via centrifugal ultrafiltration) in 19 adult males, versus a strong correlation (r = 0.858) in 20 females, underscoring its inaccuracy as a proxy in this population.[56] Due to these limitations, direct measurement of free testosterone by equilibrium dialysis or validated calculated methods incorporating albumin binding is preferred over FAI, especially when total testosterone is borderline (8–12 nmol/L).[20] 2023 UK guidelines note no established lower FAI thresholds for hypogonadism diagnosis, further diminishing its clinical utility in men.[57]In elderly men, where age-related SHBG elevations can confound total testosterone interpretations, FAI may provide supplementary insight into bioavailable androgens for symptomatic individuals with borderline results, but it is not the optimal metric. Free testosterone measurement remains the gold standard for confirming late-onset hypogonadism (andropause), as FAI overestimates or underestimates free fractions amid SHBG variability and comorbidities like obesity.[20] Studies in older men at HIV risk have utilized FAI thresholds (<14.8) to identify low androgen states associated with factors such as injection drug use and hepatitis C virus (HCV) infection, though direct free testosterone assays are recommended for precision.[58]Among children, FAI aids pubertal assessment by tracking androgen activity changes, with values rising in correlation with developmental stages, though age- and sex-specific adjustments are essential. Recent pediatric cohorts, such as the CALIPER study (2015), confirm prepubertal FAI levels are low and similar between sexes, with increases during puberty more pronounced in boys than girls, aligning with Tanner stages; however, exact ranges vary by assay and require contemporary normative data to avoid misinterpretation in delayed or precocious puberty.[22][59]In other special populations, such as those with HIV or chronic liver disease, FAI's applicability is constrained by disease-specific SHBG alterations, necessitating cautious interpretation and preference for direct free testosterone assays. In HIV-infected men, low FAI correlates with androgen deficiency risk factors like high body mass index and psychotropic medication use, yet its reliability mirrors general male limitations, with studies emphasizing free testosterone for accurate hypogonadism detection amid inflammatory effects.[58][60] For males with HCV-related chronic liver disease, lower FAI (e.g., 4.58 ± 2.10 in advanced fibrosis stage F4 vs. 7.68 ± 2.40 in F1) inversely associates with fibrosis severity, hepatic steatosis, and insulin resistance (r = -0.506 with HOMA-IR), positioning it as a useful surrogate for free testosterone decline and prognostic marker, though not superior to direct measures.[61] In transgender individuals undergoing gender-affirming hormone therapy, FAI monitoring requires adjustments for therapy-induced SHBG and testosterone shifts; for instance, drug-naïve transgender men exhibit elevated FAI relative to cisgender women, reflecting hyperandrogenism, but guidelines favor free testosterone tracking to assess metabolic impacts.[62]Emerging evidence suggests FAI's potential in cardiovascular riskstratification for men, where low values align with reduced free androgens linked to adverse outcomes, though studies prioritize direct free testosterone metrics. In older men, lower bioavailable testosterone (proxied variably by FAI) independently associates with higher cardiovascular event incidence, including heart failure mortality, prompting calls for androgen assessment in at-risk groups despite FAI's inherent male-specific inaccuracies.[63][64]