Sex hormone-binding globulin (SHBG) is a protein that binds certain androgens and estrogens with high affinity (1)(2) and restricts the access of such steroids to tissues as long as the steroids remain bound to SHBG (3). Measurements of SHBG are often done on a clinical basis and can be used to determine the amounts of free and bioavailable estradiol and testosterone (3)(4)(5). Initially, the measurements were indirect (6)(7)(8)(9), but radioimmunoassays were developed (10)(11). Now several commercial methods are available to measure SHBG directly using the interaction of SHBG with an antibody. We have found marked differences in the values obtained by these methods.
Blood samples were obtained from 20 men (ages 21–30 years) and 20 women (ages 21–42 years) between 0800 and 1000. The men and women were healthy by history and a brief physical exam, and were taking no medications. The women all had regular menstrual cycles. The samples were allowed to clot, and the serum was collected, divided into aliquots, and stored at −80 °C until analyzed. Analyses were carried out within 3 months of blood collection. The protocol was approved by the Institutional Review Board of the University of Massachusetts Medical School, and informed consent was obtained from all subjects.
Methods A and B were from Diagnostic Systems Laboratories, method C was from Radim (Angleur, Belgium through Wein Laboratories, Succasunna, NJ, and now available through KMI Diagnostics, Minneapolis, MN), and method D was from Diagnostic Products Corporation. Methods A, B, and C were radioimmunoassays using an antibody to SHBG and an antibody to γ-globulin. Method D was an immunometric assay, and method E was an indirect measurement using [3H]dihydrotestosterone and a filter-paper disc to collect the SHBG as described (7)(12).
Because the data did not follow a gaussian distribution, the analysis was done using log-transformed data. An analysis of variance indicated a significant difference between the means, and a Bonferroni correction was used to determine the significant differences between the means. A two-tailed t-test was used to analyze the differences between the means for men and women.
The characteristics of the methods tested are shown in Table 1 . Methods A–D all used an antibody to SHBG, but the antibodies were raised in different species. The intra- and interassay CVs were in the same range. A major difference was in the expected ranges for men and women, with wide variation for the maximum normal values for both men and women.
The results for the men are shown in Fig. 1 . The highest mean value was for assay B, which was approximately fourfold higher than the mean values for assays A and C and twice as high as those of D and E. The mean values for assays A and C were nearly the same (P >0.05), but both means were significantly different (P <0.05) from those for assays D and E. The values for assay B were very different from those of the other assays, whereas D and E, although differing in methodology, were not different in their results (P >0.05).
The results for the women are also shown in Fig. 1 . Similar to the results for men, the mean for method B was higher than all the others. Assays A and C were not significantly different, nor were methods D and E (P >0.05), but both assays A and C were significantly different from assays D and E (P <0.05).
When the means for men and women were compared within assays, the mean values for women were significantly (P <0.01) higher than the means for the assays in men.
The measurement of SHBG often is used clinically in investigations of hyperandrogenicity in women and hypogonadism in men. The SHBG concentration has been used, along with the testosterone concentration, to calculate the percentage and amount of non-protein, or free, testosterone circulating in the blood as well as the amount of albumin-bound testosterone (6)(13)(14). The ratio of the concentrations of testosterone to SHBG has also been used as an index of androgenicity (4)(15). Thus, the measurement of SHBG can be critical to the evaluation of certain patients. It is evident from our data that the result of these calculations can be very different depending on the method used to measure SHBG. The mean percentage of free testosterone calculated by the method of Rosner (14) varied from 1.6% for method B to 3.7% for method C, and the mean indices of androgenicity (testosterone/SHBG) were 0.29 and 0.57, respectively.
There are several reasons why the methods might give such disparate results. SHBG exists in several forms in the blood (16)(17)(18), in part because of differences in glycosylation (19)(20). Antibodies raised to SHBG may recognize these forms differently, and thus different antibodies could give disparate results in an immunoassay. Although human SHBG was always the antigen, the antibodies were raised in different species, and this could produce antibodies that recognize SHBG differently. The preparation of the antigens could differ and produce antibodies that recognize SHBG in dissimilar manners.
SHBG concentrations do differ between men and women, as our results indicate when the means for men and women were analyzed for each method separately. However, the differences between methods were the same in groups with high and low values for SHBG concentrations.
Because of the great disparity between the methods tested, it would be important to know exactly which method was used when trying to compare results reported by investigators. It would also be important to know the exact method identified by more than the source of the reagents because two of them have the same source.
Investigators must be careful in reviewing data on SHBG concentrations that the method used is clearly indicated to avoid misinterpretation.
Characteristics of the methods tested.
| Method | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| A | B | C | D | E | |||||
| Assay type | Double antibody | Coated tube | Double antibody | Immunometric | Indirect: filter paper absorbent | ||||
| Antibody | Goat anti-SHBG | −1 | Rabbit anti-SHBG | Monoclonal murine anti-SHBG | |||||
| CV | |||||||||
| Intraassay,2 % | 2.7 | 3.7 | 3.1 | 6.1 | 11 | ||||
| Interassay,2 % | 4.0 | 8.7 | 8.5 | 8.0 | 8.0 | ||||
| Range of normal,2 nmol/L | |||||||||
| Men | 10–55 | 9–111 | 13–55 | 13–71 | |||||
| Women | 30–95 | 24–220 | 30–95 | 18–114 | 11–190 | ||||
| Method | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| A | B | C | D | E | |||||
| Assay type | Double antibody | Coated tube | Double antibody | Immunometric | Indirect: filter paper absorbent | ||||
| Antibody | Goat anti-SHBG | −1 | Rabbit anti-SHBG | Monoclonal murine anti-SHBG | |||||
| CV | |||||||||
| Intraassay,2 % | 2.7 | 3.7 | 3.1 | 6.1 | 11 | ||||
| Interassay,2 % | 4.0 | 8.7 | 8.5 | 8.0 | 8.0 | ||||
| Range of normal,2 nmol/L | |||||||||
| Men | 10–55 | 9–111 | 13–55 | 13–71 | |||||
| Women | 30–95 | 24–220 | 30–95 | 18–114 | 11–190 | ||||
Manufacturer could not disclose source of antibody.
As stated by manufacturer.
Characteristics of the methods tested.
| Method | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| A | B | C | D | E | |||||
| Assay type | Double antibody | Coated tube | Double antibody | Immunometric | Indirect: filter paper absorbent | ||||
| Antibody | Goat anti-SHBG | −1 | Rabbit anti-SHBG | Monoclonal murine anti-SHBG | |||||
| CV | |||||||||
| Intraassay,2 % | 2.7 | 3.7 | 3.1 | 6.1 | 11 | ||||
| Interassay,2 % | 4.0 | 8.7 | 8.5 | 8.0 | 8.0 | ||||
| Range of normal,2 nmol/L | |||||||||
| Men | 10–55 | 9–111 | 13–55 | 13–71 | |||||
| Women | 30–95 | 24–220 | 30–95 | 18–114 | 11–190 | ||||
| Method | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| A | B | C | D | E | |||||
| Assay type | Double antibody | Coated tube | Double antibody | Immunometric | Indirect: filter paper absorbent | ||||
| Antibody | Goat anti-SHBG | −1 | Rabbit anti-SHBG | Monoclonal murine anti-SHBG | |||||
| CV | |||||||||
| Intraassay,2 % | 2.7 | 3.7 | 3.1 | 6.1 | 11 | ||||
| Interassay,2 % | 4.0 | 8.7 | 8.5 | 8.0 | 8.0 | ||||
| Range of normal,2 nmol/L | |||||||||
| Men | 10–55 | 9–111 | 13–55 | 13–71 | |||||
| Women | 30–95 | 24–220 | 30–95 | 18–114 | 11–190 | ||||
Manufacturer could not disclose source of antibody.
As stated by manufacturer.
![Mean [+ SD (bars)] SHBG concentrations (nmol/L) as measured in males and females by the respective methods.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/clinchem/46/9/10.1093_clinchem_46.9.1415/4/m_cy0902918001.gif?Expires=1674706139&Signature=zHWqs79UD5knuEPfUHQd86J83Csm7wK99zSpu1ya-QwpN~evk8BgF3sE1kL1OmmOoNE2rWK4~ZAllmTJx73EKuh8ftJeRW09TXj840iOleWTfYm91Slk94U~gy9FP4oWYmT7Xf0GgaI~9gQ7ZTw65iL2Mc15PPUv4Xi7Xad4dPfF~SFFobzp-3nHc2D6oSgpipIbaX51Vm-VZMQGHOE2vCy~0tHla23Meo0JOUHc~UA75WYA4ZdhEG5BP~mXbuDfJkry53cBuQJA3TcspW-COmIKal5i9giuqfa3WF6X~QK96Y-xJHPNfLe3f4S~wwbcxjTNtLsQQQZUYwhGllmsGg__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Mean [+ SD (bars)] SHBG concentrations (nmol/L) as measured in males and females by the respective methods.
fax 508-856-5933, e-mail christopher.longcope@umassmed.edu
References
Siiteri PK, Murai JT, Hammond GL, Nisker JA, Raymoure SJ, Kuhn RW. The serum transport of steroid hormones.
Corvol PL, Chrambach A, Rodbard D, Bardin CW. Physical properties and binding capacity of testosterone-estradiol-binding globulin in human plasma, determined by polyacrylamide gel electrophoresis.
