Biotin, which is often found at high levels in multivitamins, including supplements used for hair, skin and nail growth, can significantly interfere with certain laboratory tests and cause undetectable false results [6]. In a study performed in children and adolescents, no relationship was found between dietary biotin intake or obesity and adiposity [21]. In addition, it has been shown that there was no change in blood biotin levels before or after bariatric surgery in obese adults [22]. In our study, no significant difference was found between the control and obese groups in terms of the biotin concentration studied via the competitive method.
Normal biotin levels in blood are lower in children than in adults. The normal range of serum biotin levels in children aged 0–4 years is 4,7–22,0 ng/dl, whereas it is 8,4–20,5 ng/dl in adults [23]. Normal serum biotin concentrations in children aged 6–12 years were found as the mean ± SD (14 ± 5,01 ng/dl) with a median of 12,99 ng/dl [24]. In our study, there was a significant difference between the biotin concentrations measured by the biotin sandwich ELISA kit and the competitive ELISA kit. High biotin concentrations were observed with the sandwich ELISA kit, as shown in Table 2. The biotin values given in the above reports were evaluated. The range of normal biotin values in children's blood was compatible with the biotin values measured with the competitive kit. This finding showed that there may be severe interference in the biotin sandwich ELISA kit and that the biotin concentrations measured by the competitive method can provide more accurate results.
According to a report by the American Association for Clinical Chemistry (AACC), the results indicating biotin interference in laboratory tests included TSH and other thyroid function tests and endocrine tests such as PTH, ACTH, prolactin, testosterone and cortisol. This report also described the biotin interference thresholds or biotin-mediated interference levels of clinical tests for laboratory tests using immunologic methods by many manufacturers, such as Beckman Coulter, Inc., Roche Diagnostics, Ortho Clinical Diagnostics, and Siemens [25].
In 2016, according to the evaluation of the 8 most popular immunoassay analyzers used in the United States, it was shown that exogenous biotin in the sample can cause more than ± 10% biotin interference in test results from sandwich and competitive methods [26].
In another study, it was shown that high-dose biotin caused false high results in cortisol, cyclosporine A, digitoxin, tacrolimus, FT3 and FT4 competitive immunological tests and false low results in C-peptide, insulin, TSH, NT-proBNP, high-sensitivity troponin T, HIV, procalcitonin and b-HCG sandwich immunological tests [27]. False negative results in troponin T tests are vital for patients, and newer cardiac troponin T test formulations are now being sought to have a much higher biotin interference threshold [28]. In a study of acute myocardial infarction patients, the probability of biotin interaction giving false negative results to the Elecsys Troponin T Gen 5 (TnT Gen 5; marketed outside the United States [US] as Elecsys Troponin T-high sensitive; Roche Diagnostics International Ltd., Rotkreuz, Switzerland) test was found to be very low [29]. Dilution and biotin removal procedures could be used to reduce biotin interference in immunoassays [30].
Bonferroni correction and the Mann‒Whitney U test were applied to evaluate the subgroups, as shown in Table 4. According to the biotin sandwich ELISA results, biotin levels were generally greater in the healthy group than in the obese group, while in the 5th group, biotin levels were greater in the obese group than in the healthy group. With respect to the biotin competitive parameter, in some subgroups, the value in the healthy group was greater than that in the obese group, whereas in the other groups, the value was greater than that in the healthy group. There was an apparent discrepancy. This finding suggested that biotin interference in the kits may lead to erroneous results. Kabiri et al. demonstrated that biotin interference in automated systems can lead to both false low and false high results in the sandwich method [31]. Although it was generally stated that there was a positive interaction in competitive tests and a negative interaction in sandwich immunoassays [27, 32], our study showed that both false low and false high results can be obtained with both principles in manual ELISA kits.
Dasgupta et al. demonstrated that in the Cobas E-411 (Roche Diagnostics) immunoassay IL-6 test, maximum negative interference was observed with 100 ng/ml biotin, but no interference occurred with 1000 ng/ml biotin [33]. Knudsen et al. demonstrated that aldosterone and renin analyses performed in automated systems showed a very high rate of interference with biotin at doses of 100–500 ng/ml, while insulin-like growth factor 1, growth hormone and bone alkaline phosphatase analyses were less sensitive to biotin doses in this range [34]. Choi et al. showed that biotin interference is still present in second- and third-generation Elecsys (Roche Diagnostics) free thyroxine immunoassay tests updated by the company [35]. Biotin levels are not measured in all laboratories, which makes the problem of biotin interference difficult to solve. Dilution tests and neutralization analyses may help to solve this problem [36].
Although it varies according to the type of immunological test used in automated systems, it has been reported that it is possible to use a dose of 10 ng/ml in blood as the lowest threshold value for biotin interference [25, 37]. For all the ELISA parameters, biotin % bias values between the groups formed by the addition of biotin at concentrations of 10, 50, 500, and 1000 ng/ml were calculated [4, 29, 31]. In our study, there was biotin interference in all ELISA parameters commonly used in investigations performed in the field of pediatric obesity, and this interference was found to be greater in the healthy group. In addition, the addition of 50, 500, or 1000 ng/ml biotin to the studied parameters created both negative and positive interferences. The % bias values for all parameters ranged between 1,4-113,6% for the healthy group and 0,17–49,2% for the obese group. For example, for all three biotin concentrations of the leptin kit, the % bias value in both the healthy and obese groups was < 5,6, the % bias value of the GHSR kit in the healthy group was < 5,7, the % bias value of the LEPR kit in the healthy group was < 11,6, and that in the obese group was < 3,4. Specifically, biotin interference is closely related to the quality of the ELISA kit, the parameters studied and the concentration of biotin in the blood.