Transferrin
Clinical utility
Transferrin is the plasma protein that binds and transports ferric iron to cells, keeping iron non-reactive in circulation. Its concentration (or the related total iron-binding capacity, TIBC) is measured together with serum iron to calculate transferrin saturation (TSAT), which is used to assess iron deficiency, iron overload (e.g., hereditary hemochromatosis), and to help distinguish iron deficiency anemia from anemia of chronic disease. Transferrin is decreased in iron overload, chronic liver disease, malnutrition, and nephrotic syndrome, and increased in iron deficiency; because it is a negative acute-phase protein it is also decreased in chronic inflammatory disorders and malignancy.
How it's measured
Direct immunologic measurement of transferrin
Transferrin can be measured directly by immunologic techniques. The first immunodiffusion method for transferrin was described in 1965; this was largely replaced by automated immunoturbidimetric or immunonephelometric procedures developed in the 1970s, which remain the basis of modern automated transferrin assays.
Indirect measurement via iron-binding capacity (TIBC/UIBC)
Transferrin can alternatively be quantified indirectly in terms of the amount of iron it will bind (total iron-binding capacity, TIBC), a practical/chemical method first reported in 1957 and later standardized by the International Committee for Standardization in Haematology (1978, revised 1990). This three-step procedure can be written as:
Unbound excess Fe3+ + solid adsorbent (magnesium carbonate, charcoal, or ion-exchange resin) --adsorption--> Fe3+ removed from solution (bound to adsorbent) + iron-saturated transferrin remaining in solution
Iron-saturated transferrin --acidic pH--> apotransferrin + free Fe3+
Free Fe3+ + chromogen (e.g., bathophenanthroline, ferrine, or ferrozine) --acidic pH--> colored Fe-chromogen complex (measured colorimetrically)
In the 1990s more direct TIBC assays were developed, and most automated chemistry analyzers instead measure unsaturated iron-binding capacity (UIBC) and calculate TIBC (TIBC = UIBC + serum iron), because this is more easily automated.
Transferrin (g/L) and TIBC can be interconverted using derived formulas (e.g., Transferrin (g/L) = 0.007 × TIBC (µg/L)), though the source notes these mathematical derivations have recognized weaknesses (nonspecific iron binding to albumin in the TIBC assay, variability in transferrin molecular mass due to glycation, and disease-related shifts in the TIBC-transferrin relationship), and manufacturers often use an experimentally derived conversion factor (1.27) rather than the theoretical one (1.41). Since the introduction of an international protein reference material (CRM 470/ERM-DA470) in 1994, direct immunochemical transferrin measurement (rather than TIBC) has been recommended because of better reproducibility and standardization; TIBC may still be preferred in non-European populations with genetic transferrin variants (BC/CD), where immunochemical transferrin determination may be less accurate.
Typical reference interval
| Population | Reference interval |
|---|---|
| General (interim consensus, IFCC/CRM 470-standardized, 1996) | 2.0–3.6 g/L |
| Japanese population (same standardization) | 1.9–3.2 g/L |
The source notes these values are similar, suggesting ethnic differences in transferrin reference intervals are not very pronounced.
Sources: Ch40 Iron metabolism.pdf · Extracted 2026-08-17 · Approved 2026-08-20
Method comparison
Beckman Coulter AU5800
kinetics endpoint
AMR 75 – 750 mg/dL
Immunoturbidimetric endpoint assay; complex formation increases light scatter, measured as increased absorbance (decreased light transmittance), not a catalyzed chemical reaction.
- Reaction notes:
This is an immunoturbidimetric method, not a catalyzed chemical reaction, so there is no limiting-reagent/rate-limiting-step chemistry to describe from the insert; the signal actually monitored is the increase in absorbance (decrease in transmitted light) caused by light scattering off the immune complexes as they form in the Tris-buffered, PEG-containing reagent.
- Kinetics: The analyzer reads all determinations at the same fixed time interval (an endpoint-style read) rather than tracking a continuous rate; the procedure is linear over the analytical measuring range of 75–750 mg/dL, and samples above 750 mg/dL must be diluted and repeated.
- Readout: Turbidimetric measurement: the reduction in light transmitted (equivalently, the increase in absorbance) through the reaction mixture, caused by antigen-antibody complexes scattering light in proportion to their size, shape, and concentration, is read and converted to a transferrin concentration.
- What's measured: The assay measures total serum transferrin protein via its immunological reaction with goat anti-transferrin antiserum; it does not report iron-saturation or iron-binding capacity, only the transferrin protein concentration.
- Interference implications: Bilirubin, hemolysis, and lipemia (simulated with Intralipid) were each tested and found not to significantly interfere at the stated levels (bilirubin up to 40 mg/dL, hemolysate up to 500 mg/dL, Intralipid up to 1,000 mg/dL), using a criterion of recovery within 10% of the initial value. No interferents with a confirmed quantified bias are listed.
- Known bias: Method comparison of DxC 700 AU (y) vs. AU5800 (x) on patient serum samples (n=123, range 85–647 mg/dL) by Deming regression gave slope 0.987, intercept -3.4, and correlation coefficient r=0.9994, indicating close agreement between the two platforms.
Interferences & pre-/post-analytical notes
Interferences
Validated -- no significant interference
- Bilirubin: No significant interference (up to 40 mg/dL bilirubin) — CLSI EP07-A2, Interference Testing in Clinical Chemistry (reference 4)
- Hemolysis (hemolysate): No significant interference (up to 500 mg/dL hemolysate) — CLSI EP07-A2, Interference Testing in Clinical Chemistry (reference 4)
- Lipemia (Intralipid): No significant interference (up to 1,000 mg/dL Intralipid) — CLSI EP07-A2, Interference Testing in Clinical Chemistry (reference 4)
Pre-analytical
- Fasting serum specimen, free from hemolysis, is the recommended specimen.
- Transferrin is stable at ≤ -20°C for over 40 days and can be repeatedly frozen and thawed without loss of immunological activity.
- Opened reagent bottles are stable for 90 days when stored in the refrigerated compartment of the analyzer; unopened reagents are stable until the expiration date when stored at 2–8°C.
Post-analytical
- Samples exceeding the upper limit of linearity (750 mg/dL) should be diluted and repeated; the AUTO REPEAT RUN feature can dilute, repeat, and multiply by the dilution factor automatically.
- Default unit of measure is mg/dL; for conversion to SI units (g/L), multiply the result by 0.01.
Roche Cobas c503
λ 505 / 700 nm
AMR 0.1-5.2 g/L (1.26-65.5 μmol/L, 10-520 mg/dL)
LoD 0.1 g/L (1.26 μmol/L, 10 mg/dL)
LoQ 0.1 g/L (1.26 μmol/L, 10 mg/dL)
Immunoturbidimetric assay: antibody-mediated agglutination of transferrin, read as an increase in turbidity rather than a colored enzymatic product.
- Reaction notes:
This is antibody-antigen agglutination, not an enzymatic reaction: antibody titer (R3) is set by the manufacturer to keep the reaction within the reportable range, and the signal monitored is light scattering/absorbance from the growing immune complex, not a discrete chemical product. The insert does not specify which reagent is in excess or the reaction's rate characteristics.
- Readout: Turbidity generated by antigen-antibody complex formation is measured bichromatically at 505 nm (main) and 700 nm (sub); the instrument converts the signal to transferrin concentration in g/L (also displayable in μmol/L and mg/dL via conversion factors).
- What's measured: Measures total transferrin protein by immunoturbidimetric agglutination against human transferrin's polypeptide/glycan structure; does not distinguish iron-saturated from unsaturated transferrin — saturation must be derived separately (e.g., with serum iron).
- Interference implications: Icterus, hemolysis, lipemia, rheumatoid factor, and common therapeutic drugs were all tested and shown not to interfere significantly at the stated indices/concentrations, and no high-dose hook effect occurs up to 17 g/L transferrin. The only flagged risk is gammopathy (especially IgM/Waldenström's macroglobulinemia), which the insert says may rarely cause unreliable results.
- Known bias: Insert reports only inter-instrument correlation using the same reagent (cobas c503, c303, and c703 vs. cobas c501), not a comparison to a different method: Passing/Bablok slopes ranged ~0.96-1.01 with r ≥0.997, indicating no significant bias between these analyzer platforms.
Interferences & pre-/post-analytical notes
Interferences
Suspected / theoretical
- Gammopathy, particularly type IgM (Waldenström's macroglobulinemia): May cause unreliable (unspecified direction) results (in very rare cases) — Bakker AJ, Mücke M. Gammopathy interference in clinical chemistry assays: mechanisms, detection and prevention. Clin Chem Lab Med 2007;45(9):1240-1243.
Validated -- no significant interference
- Icterus (conjugated and unconjugated bilirubin): No significant interference (up to an I index of 60 (approximate conjugated and unconjugated bilirubin concentration: 1026 μmol/L or 60 mg/dL)) — Glick MR, Ryder KW, Jackson SA. Graphical Comparisons of Interferences in Clinical Chemistry Instrumentation. Clin Chem 1986;32:470-475.
- Hemolysis (hemoglobin): No significant interference (up to an H index of 1000 (approximate hemoglobin concentration: 621 μmol/L or 1000 mg/dL)) — Glick MR, Ryder KW, Jackson SA. Graphical Comparisons of Interferences in Clinical Chemistry Instrumentation. Clin Chem 1986;32:470-475.
- Lipemia (Intralipid): No significant interference; poor correlation between L index (turbidity) and triglyceride concentration (up to an L index of 500) — Glick MR, Ryder KW, Jackson SA. Graphical Comparisons of Interferences in Clinical Chemistry Instrumentation. Clin Chem 1986;32:470-475.
- Rheumatoid factors: No significant interference (up to a concentration of 1200 IU/mL)
- High dose hook effect (transferrin): No false result occurs (up to a transferrin concentration of 17 g/L)
- Drugs (common therapeutic drug panels): No interference found at therapeutic concentrations (therapeutic concentrations per standard drug interference panels) — Breuer J. Report on the Symposium "Drug effects in Clinical Chemistry Methods". Eur J Clin Chem Clin Biochem 1996;34:385-386; Sonntag O, Scholer A. Drug interference in clinical chemistry. Ann Clin Biochem 2001;38:376-385.
Pre-analytical
- Only serum and Li-heparin plasma were tested and found acceptable; do not use EDTA or citrate plasma.
- Centrifuge samples containing precipitates before performing the assay.
- Sample stability: 8 days at 15-25°C, 8 days at 2-8°C, 6 months at (-15)-(-25)°C; freeze only once.
Post-analytical
- Samples exceeding the 5.2 g/L measuring range should be run via the rerun function, which applies a 1:1.5 sample dilution; results are automatically multiplied by a factor of 1.5.
- cobas c systems report transferrin in g/L, with conversion factors of ×12.6 for μmol/L and ×100 for mg/dL.
Roche Cobas c503 — Urine
kinetics endpoint
λ 340 / 700 nm
AMR 2.2-35.0 mg/L (0.22-3.5 mg/dL)
LoD 1.5 mg/L (0.15 mg/dL)
LoQ 2.2 mg/L (0.22 mg/dL)
Immunoturbidimetric assay (no stoichiometric reaction equation stated): human transferrin in the urine sample reacts with anti-human transferrin antibodies (rabbit, R3) to form an antigen-antibody precipitate/immune complex, which is quantified by measuring the resulting turbidity (light scatter/absorbance) bichromatically at 700/340 nm.
- Reaction notes:
No stoichiometric reaction equation is stated — this is an immunoturbidimetric (antigen-antibody agglutination) assay: anti-human transferrin antibody (rabbit, reagent R3) binds transferrin antigen in the urine sample to form an insoluble immune complex, and the resulting turbidity is measured photometrically at 700/340 nm (bichromatic). Antibody amount is described only as 'dependent on titer'; the insert gives no concentration data to derive an excess fold. The signal monitored is light scatter/absorbance from immune complex formation, not transferrin directly.
- Kinetics: Reported as a single reading after a fixed 10-minute reporting time with a non-linear calibration curve (S1-S6), consistent with a fixed-time/endpoint immunoturbidimetric measurement rather than a continuously monitored rate reaction. The measuring range is narrow (2.2-35.0 mg/L); samples above range require a 1:3 rerun dilution with automatic 3x result multiplication.
- Readout: Turbidity generated by transferrin-antibody immune complex formation is read bichromatically (700 nm sub/340 nm main); increasing turbidity corresponds to increasing urinary transferrin concentration.
- What's measured: Measures total immunoreactive transferrin in urine (not iron-bound/saturation-specific). Intended for quantitation of urinary transferrin, which together with albumin results allows estimation of the charge selectivity of glomerular defects, since the two proteins are similar in size but differ in charge.
- Interference implications: A broad panel of urine constituents (albumin, calcium, citrate, creatinine, glucose, IgG, magnesium, oxalate, phosphate, urea, uric acid, urobilinogen) and hemolysis up to H index 20 show no significant interference at the stated recovery criterion (±0.5 mg/L at ≤5.0 mg/L, ±10% above). No high-dose hook effect up to 650 mg/L transferrin. Common therapeutic drug panels show no interference except ofloxacin, which causes falsely elevated results. Rare IgM gammopathy (Waldenström's macroglobulinemia) may cause unreliable results.
- Known bias: The insert reports platform-to-platform comparisons of the same reagent (not a comparison to a different assay method): cobas c 503 vs c501, Passing/Bablok y=1.028x-0.170 mg/L (r=0.998, n=74); cobas c 303 vs c501, y=1.065x-0.0249 mg/L (r=0.996, n=60); cobas c 703 vs c503, y=0.987x-0.0967 mg/L (r=0.999, n=62).
Interferences & pre-/post-analytical notes
Interferences
Known interferences
- Ofloxacine: Causes artificially high (falsely elevated) transferrin results — Sonntag O, Scholer A. Drug interference in clinical chemistry: recommendation of drugs and their concentrations to be used in drug interference studies. Ann Clin Biochem 2001;38:376-385.
Suspected / theoretical
- Gammopathy, particularly type IgM (Waldenström's macroglobulinemia): May cause unreliable results (In very rare cases) — Bakker AJ, Mücke M. Gammopathy interference in clinical chemistry assays: mechanisms, detection and prevention. Clin Chem Lab Med 2007;45(9):1240-1243.
Validated -- no significant interference
- Hemolysis (H index): No significant interference; urine specimens colored due to presence of erythrocytes or hemoglobin should not be used (H index up to 20) — Glick MR, Ryder KW, Jackson SA. Graphical Comparisons of Interferences in Clinical Chemistry Instrumentation. Clin Chem 1986;32:470-475.
- Albumin: No significant interference (≤ 5000 mg/L)
- Calcium: No significant interference (≤ 8 mmol/L (32 mg/dL))
- Citrate: No significant interference (≤ 10 mmol/L)
- Creatinine: No significant interference (≤ 44 mmol/L (497 mg/dL))
- Glucose: No significant interference (≤ 111 mmol/L (2000 mg/dL))
- IgG: No significant interference (≤ 500 mg/L)
- Magnesium: No significant interference (≤ 75 mmol/L (182 mg/dL))
- Oxalate: No significant interference (≤ 2.2 mmol/L)
- Phosphate: No significant interference (≤ 40 mmol/L (124 mg/dL))
- Urea: No significant interference (≤ 1000 mmol/L)
- Uric acid: No significant interference (≤ 6 mmol/L (101 mg/dL))
- Urobilinogen: No significant interference (≤ 15 mg/dL)
- High-dose hook effect (transferrin): No false result occurs (Up to a transferrin concentration of 650 mg/L (65 mg/dL))
- Drugs (common therapeutic panel): No interference found at therapeutic concentrations
Pre-analytical
- Only urine specimens were tested and found acceptable for this application; each urine sample must be centrifuged (10 minutes at approximately 3000 x g) prior to testing.
- Random and timed urine collections are suitable specimens for testing transferrin in urine.
- The pH of the urine should be adjusted to pH 7.0.
- Urine specimens colored due to presence of erythrocytes or hemoglobin should not be used.
- Stability in urine: 4 days at 15-25°C; 7 days at 2-8°C. Freezing and thawing is not allowed.
Post-analytical
- Samples with concentrations above the measuring range (35.0 mg/L) should be determined via the rerun function, which performs a 1:3 dilution and automatically multiplies results by a factor of 3.
- Conversion factor: mg/L × 0.1 = mg/dL.
- Expected value: transferrin in urine of healthy individuals is ≤ 1.9 mg/L (0.19 mg/dL), which is below the detection limit of this method; each laboratory should establish its own reference ranges.
- Special wash programming is mandatory when certain test combinations are run together on cobas c systems, to avoid carry-over.
Siemens Atellica CH
kinetics endpoint
λ 340 / 596 nm
AMR 1–440 mg/dL (0.01–4.40 g/L)
LoD 1 mg/dL (0.01 g/L)
PEG-enhanced immunoturbidimetric assay (Heidelberger-Kendall precipitin reaction as modified by Hellsing); turbidity from the immune complex is measured bichromatically and compared to a standard curve built from calibrator absorbances.
- Reaction notes:
PEG (6%) enhances aggregation of antigen-antibody complexes into a measurable precipitate; the antibody reagent (R2, 17 μL of 5.0 mL antiserum) is added after an initial sample-only absorbance reading, so the reaction is effectively a fixed-time endpoint rather than a monitored rate. The insert does not give antibody concentration, so the exact molar excess over transferrin cannot be derived, but reagent volumes (80 μL R1 + 17 μL R2) greatly exceed the 2 μL of prediluted sample. The signal monitored is turbidity/light scattering from the immune complex, not transferrin itself.
- Kinetics: Absorbance is read once before and once after antibody addition (fixed-time/endpoint design), and concentration is read off a calibrator-derived standard curve. This underlies the linear reportable range of 1–440 mg/dL; values above this trigger an optional autorepeat with dilution to extend reporting up to 4400 mg/dL.
- Readout: Bichromatic absorbance (340/596 nm) is measured before and after the antibody reagent is added and the mixture incubated at 37°C; the absorbance change reflects the turbidity generated by immune complex formation and is converted to a transferrin concentration via the calibration curve.
- What's measured: Measures total transferrin in serum or plasma (lithium heparin, potassium EDTA only); the insert does not distinguish iron-saturation forms or glycoforms. Correlates with total iron binding capacity (TIBC) per the Summary and Explanation section.
- Interference implications: Hemolysis (Hb up to 525 mg/dL), conjugated and unconjugated bilirubin (up to 30 mg/dL each), and lipemia/Intralipid (up to 650 mg/dL) were all tested per CLSI EP07-A2 and produced biases well within the assay's ≤10% interference design goal (range -7% to 0%), so none are considered significant interferents at the levels tested. Separately, very high transferrin concentrations (>7200 mg/dL) cause a high-dose hook effect, reading as falsely low (>440 mg/dL reported as the ceiling) rather than reflecting the true elevated value.
- Known bias: Versus ADVIA Chemistry 1800 TRF (serum, N=102, range 4–422 mg/dL): y = 0.98x - 3 mg/dL (y = 0.98x - 0.03 g/L), r = 0.997, indicating close agreement with a slight negative bias at the low end of the equation.
Interferences & pre-/post-analytical notes
Interferences
Validated -- no significant interference
- Hemoglobin: Bias of -1% to 0% at tested transferrin concentrations of 121 and 281 mg/dL; assay designed for ≤10% interference and bias >10% is considered interference (525 mg/dL (0.326 mmol/L) hemoglobin)
- Bilirubin, conjugated: Bias of -1% to 0% at tested transferrin concentrations of 119 and 285 mg/dL (30 mg/dL (513 μmol/L) conjugated bilirubin)
- Bilirubin, unconjugated: Bias of -1% at tested transferrin concentrations of 121 and 290 mg/dL (30 mg/dL (513 μmol/L) unconjugated bilirubin)
- Lipemia (Intralipid): Bias of -7% at tested transferrin concentrations of 119 and 285 mg/dL; within the assay's ≤10% interference design goal (650 mg/dL (7.35 mmol/L) Intralipid)
Pre-analytical
- Specimen types: serum and plasma (lithium heparin, potassium EDTA); sample volume 2 μL per determination.
- Separated specimens are stable up to 8 hours at room temperature, up to 7 days at 4–8°C, or up to 1 month frozen at -20°C.
- Samples must be free of bubbles/foam and fibrin/particulate matter before analysis; remove particulates by centrifugation per CLSI guidance.
Post-analytical
- System flags all values outside the 1–440 mg/dL (0.01–4.40 g/L) measuring interval; an optional automatic repeat (with dilution) extends the reportable range to 4400 mg/dL (44.00 g/L), flagged 'Autorepeat'.
- High-dose hook effect: transferrin concentrations as high as 7200 mg/dL (72.00 g/L) will read as >440 mg/dL (>4.40 g/L) rather than falsely low, due to a paradoxical decrease in signal at very high analyte levels.
Comparison highlights
- All four assays measure transferrin directly via immunoturbidimetry, consistent with the overview's note that direct immunochemical measurement (rather than indirect TIBC/iron-binding methods) is now the recommended approach; none of these four reports TIBC or iron-saturation, only total transferrin protein.
- Siemens Atellica CH uses PEG-enhanced immunoturbidimetry (Heidelberger-Kendall precipitin reaction as modified by Hellsing), an enhancement chemistry not described for the Beckman or Roche assays, which use plain antibody-mediated agglutination/immune-complex turbidity.
- Detection wavelengths differ substantially across platforms: Roche c503 serum reads 505/700 nm, the Roche c503 urine assay 340/700 nm, and Siemens Atellica CH 340/596 nm (Beckman's wavelength isn't stated) — despite all resting on the same antigen-antibody turbidity principle.
- High-dose hook effect behavior differs: Roche's serum and urine assays were validated hook-free up to 17 g/L and 650 mg/L transferrin respectively, whereas Siemens Atellica CH does exhibit a hook effect above 7200 mg/dL that reads out falsely LOW rather than flagging as abnormal.
- Only Roche flags IgM gammopathy as a risk, on both its serum and urine assays (rare, Waldenström's macroglobulinemia can give unreliable results); Beckman's and Siemens's interference panels don't mention paraprotein/gammopathy testing at all.
- Roche's urine assay uniquely lists ofloxacin as an interferent, causing falsely elevated transferrin — a drug effect not reported for any of the serum-based methods, relevant if a urine sample may contain a fluoroquinolone.
- Cross-platform bias data exists only for Beckman and Siemens: Beckman's DxC700 AU vs AU5800 comparison (slope 0.987) and Siemens's vs ADVIA 1800 TRF comparison (slope 0.98) both show close agreement with slight negative bias. Roche's reported comparisons (c503/c303/c703 vs c501) are inter-instrument with the identical reagent, so they don't demonstrate agreement with a different manufacturer's method.
- Siemens reports quantified interference bias (e.g., -7% for lipemia, -1% to 0% for hemolysis/bilirubin) rather than the simple pass/fail "no significant interference" calls used by Beckman and Roche, giving finer-grained data for judging acceptability against a lab's own bias tolerance.
Generated 2026-08-20 · Approved 2026-08-20
Lab comparison
No lab offerings recorded for this analyte yet.