Urinary Bisphenols as Biomarkers of Exposure to Bisphenol A, Bisphenol F, and Bisphenol S: A New Procedure for Biomonitoring

Abstract

Introduction. Bisphenols are chemicals widely used in industrial production, but they are also of significant concern due to their potential health effects. In particular, bisphenol A (BPA) is a documented endocrine disruptor. As a result, bisphenol S (BPS) and bisphenol F (BPF) are now frequently used as substitutes. However, evidence of endocrine-disrupting properties is also emerging for these substances. Methods. A new, simple, and rapid HPLC-MS/MS procedure for the urinary analysis of bisphenols was developed and validated. Results. The mean accuracy ranged from 83.3 to 119.2%, and precision values (CV%) ranged from 2.2 to 18.9%. The limit of detections (LODs) for BPA, BPS, and BPF, were 0.01, 0.001, and 0.07 µg/L. The pre-analytical step involved only enzymatic hydrolysis, followed by a liquid–liquid extraction, for the subsequent urine analysis of the three BPs. Chromatographic separation was achieved in 9 min, with high selectivity. Discussion. The procedure was applied to 36 urine samples of a male population attending a fertility center. Most of the subjects showed relevant exposure to BPs (BPS average level: 0.87 ± 3.03 µg/g creatinine; BPF average level: 0.14 ± 0.34 µg/g creatinine), particularly to BPA (average level 0.97 ± 1.27 µg/g creatinine). The procedure demonstrated high efficiency and was confirmed to be practical, fast, and accurate.

1. Introduction

Bisphenols (BPs) are molecules consisting of two phenol groups linked by an external atom that may also carry other functional groups. They are involved in many industrial processes. In particular, bisphenol A (BPA) has been widely adopted in various industrial cycles. It has been extensively used [1] in products such as cans, plastic coating applied to teeth, food packaging, and personal care products. Polycarbonate plastic and epoxy resins are also derived from this base chemical.

BPA is an endocrine disruptor and a reprotoxic substance. Numerous epidemiological studies have shown adverse health effects resulting from exposure, particularly concerning male fertility [1]. This evidence has led to a gradual reduction in the use of BPA in Europe, with specific legal restrictions, and its inclusion in the authorization list under the REACH Regulation for its endocrine-disrupting properties [2].

Among the various chemicals used as alternatives to BPA, bisphenol S (BPS) and bisphenol F (BPF) are the most common. Their molecular structures are shown in Figure 1. They have relatively short blood half-lives (<6 h) and do not persist in the environment, but their potential interference with the endocrine system remains a topic of scientific interest and concern [3].

The metabolism of BPA is well known. It is primarily metabolized through glucuronidation and sulfation reactions to enhance water solubility and is mainly excreted in urine, with a half-life of less than 6 h [3]. In particular, total urinary BPA consisted of 71% of BPA glucuronide, 15% BPA sulfate, and 14% of free BPA [4]. Glucuronidation is also the main metabolic pathway for BPS, with BPS glucuronide accounting for 50.2% of total BPS. A significantly higher proportion of sulfated BPA in males compared to females was observed, while no gender-specific patterns were observed for BPS. However, males exhibited a significantly higher proportion of free BPS than females and the ratio of free BPS to total BPS was considerably greater than the proportion of free BPA to total BPA. These results suggest a sex-dependent metabolism of BPA and BPS in humans, and there are discrepancies in the metabolism of BPA and BPS in humans [5].

The metabolism of the BPF has not been extensively studied. However, it has been suggested that it could be considered similar to BPA, given their comparable half-lives. For these reasons, the most suitable biomarkers for evaluating recent exposure to BPs are the urinary unconjugated BPs, obtained by enzymatic hydrolysis [6].

Urinary BPs analysis presents certain operational challenges, primarily related to the potential contamination of the samples from plastic materials or laboratory reagents, containing BPA [7]. Specifically, the use of solid phase extraction (SPE) cartridges, may pose a contamination risk. To mitigate this issue, it is necessary to use high quality solvents, glass materials, and carefully define analytical protocols that include periodic checks with blank samples.

Several studies have been published regarding the analysis of urinary bisphenols; in particular, different analytical approaches have been proposed to measure unconjugated and total BPA [7,8,9,10,11]. Contamination by BPA trace concentrations is so pervasive during the sample pretreatment with multiple steps that in almost all the published papers, even with care, it resulted in being extremely difficult to completely exclude the introduction of BPA into biological samples. This situation commonly results in a higher signal-to-noise ratio and generally higher detection limits [12,13,14]. In consideration of the health effects that these substances can have even at low doses, in particular BPA [15], being able to have methods with low detection limits represents an undoubted advantage in biomonitoring.

This study proposes a new, rapid, and simple analytical procedure, using HPLC-MS/MS for the determination of urinary BPs. Specifically, the pre-analytical step involves only enzymatic hydrolysis, followed by liquid–liquid extraction for the subsequent analysis of the three BPs in urine.

2. Materials and Methods

2.1. Chemicals and Supplies

Pure standards of BPA, BPS, BPF, and the deuterated internal standard (ISTD) BPAD6 were supplied by Sigma Aldrich (Saint Louis, MO, USA). The β-glucuronidase-arylsulfatase enzyme from Elix Pomatia was provided by Roche (Mannheim, Germany). An Anotop 10LC syringe filter device (0.2 µm pore size, 10 mm diameter) was purchased from Whatman Inc. (Maidstone, UK). n-Hexane was supplied from Fluka (Sigma Aldrich, Germany), methanol from Carlo Erba Reagents s.r.l (Milan, Italy), ethyl acetate from Riedel-da Haen (Buchs, Switzerland), glacial acetic acid from Sigma Aldrich (Saint Louis, MO, USA), and purified water from a Milli-Q Plus System (Millipore Milford, MA, USA). Samples were analyzed on a HPLC EXIONLC AC system, supplied by Sciex, coupled with an SCIEX API 4000 triple quadrupole mass spectrometry detector equipped with a Turbo Ion Spray (TIS) probe (SCIEX, Warrington, Cheshire, UK). A Phenomenex Luna C8, 250 × 4.6 mm, 100 Å particle size, reverse phase, chromatographic column was used, thermostated at 40 °C. Glass vials (11.6 × 32 mm wide mouth) for analysis were purchased from Waters; blank urine samples were obtained from non-exposed volunteers.

2.2. Analytical Procedure

The mobile phase consisted of a gradient of water/acetonitrile. The flow rate was set at 1.0 mL/min. The solvents gradient started at 50/50 acetonitrile/water, was held in isocratic mode for 1 min, followed by a linear gradient to 100% acetonitrile in 6 min, isocratic 100% acetonitrile for 2 min, and returned to the initial conditions in 2 min, for a total analysis time of 9 min. The injection volume was 10 µL.

The mass spectrometer was set with an electrospray ionization in negative ion mode.

The mass on charge (m/z) ion transitions were monitored in negative ion mode and are presented in

Table 1. Validation and instrumental parameters.

Analyte	Use	Precurson Ion (m/z)	Ion Species	Product Ion (m/z)	Ion Species	DP 1 (V)	EP 2 (V)	CE 3 (eV)	CXP 4 (eV)
BPA	quantitative	227.0	[M-H]−	211.1	[M-H-CH4]−	−80	−15	−32	−5
BPS	qualitative	227.0	[M-H]−	133.0	[M-H-C6H6O]−	−70	−10	−26	−13
	quantitative	248.9	[M-H]−	108.2	[M-H-C6H5O-SO]−	−80	−10	−28	−11
BPF	qualitative	248.9	[M-H]−	185.0	[M-H-SO2]−	−90	−10	−36	−6
	quantitative	199.0	[M-H]−	93.0	[M-H-C7H6O]−	−80	−15	−26	−11
BPAD6	qualitative	199.0	[M-H]−	123.0	[M-H-C6H6O+H2O]−	−70	−10	−32	−5
	quantitative	233.0	[M-H]−	138.2	[M-H-C6H6O]−	−60	−15	−30	−15
	CAD 5 (eV)	CUR 6 (psi)	GS1 7 (psi)	GS2 8 (psi)	IS 9 (V)	T 10 (°C)
	10	18	45	25	−4500	400
	Retention Time (min)	LOD 11 (ng/mL)	Concentration (ng/mL)	Precision Interday (CV 12 %)	Precision Intraday (CV%)	Accuracy Interday (%)	Accuracy Intraday (%)	Mean Recovery (%)
BPA	8.1	0.01	2	18.9	3.6	101.4	83.3	92.4
			5	9.3	5.4	112.0	102.6	107.3
			20	8.5	6.1	113.5	119.2	116.3
BPS	5.7	0.001	2	8.9	7.2	115.1	106.7	110.9
			5	6.2	6.0	108.4	106.0	107.2
			20	5.6	2.2	106.8	105.7	106.3
BPF	7.3	0.07	2	12.6	7.5	96.8	85.8	91.3
			5	7.0	2.4	104.9	97.1	101.0
			20	6.1	5.5	112.6	116.2	114.4
BPAD6	8.05

¹ DE—declustering potential; ² EP—entrance potential; ³ CE—collision energy; ⁴ CXP—collision energy exit potential; ⁵ CAD—collision activated dissociation; ⁶ CUR—curtain gas; ⁷ GS1—nebulizer gas; ⁸ GS2—heater gas; ⁹ IS—interface voltage; ¹⁰ T—temperature; ¹¹ LOD—limit of detection; ¹² CV—coefficient of variation.

, together with the mass spectrometry parameters. We used BPA-D6 as the only internal standard, due to the difficulty in obtaining other marked standards, and given the very similar chemical structure, the use of BPA-D6 as an internal standard was sufficient for our purposes.

Validation of the procedure was conducted using quality controls (QCs) and prepared by adding pure standards to “blank urines” at three levels of concentration: 2 µg/L, 5 µg/L, and 20 µg/L. Blank urines were supplied by healthy volunteers, were previously analyzed for BPs, and showed no contamination.

The QCs were analyzed in triplicate, over three non-consecutives days, and the results were obtained using a calibration curve in urine. Lower limit of quantification (LOQ) was calculated using the Analyst software 1.5 supplied with the mass spectrometer, based on the standard deviation of the background noise around the analytical peak, and the LOQ was set as the concentration in µg/L at least 6 times greater than the standard deviation of noise. Otherwise, the LOD was set as the concentration in µg/L at least 3 times greater than the standard deviation of noise [16].

Spot urine samples were collected in polypropylene sterile urine containers and stored at −25 °C until analysis. The possible contamination from BPA of plastics was avoided with a preliminary analysis.

All phases were conducted to minimize accidental contamination from laboratory plastics. Glass containers, test tubes, and pipettes were used whenever possible, and when plastic was unavoidable, it was tested to exclude the possible BPA release under working conditions. Notably, before using each solvent, an analysis was conducted to see if even minimal contamination was present; the distilled water was obtained only with glass device. For the used plastic material, the contact with the solvents was evaluated, for the time and conditions foreseen by the analytical procedure, and the analysis was carried out to exclude possible contamination. Also, blank samples were previously analyzed for possible traces of bisphenol A, from the environment and from laboratory supplies, resulting in no-noticeable contamination.

Samples (5 mL) were mixed with 50 µL of β-glucuronidase-arylsulfatase enzyme from Elix Pomatia, 1 mL of 2% (v/v) acetic acid, and the deuterium labeled internal standard (BPA-D6), then incubated at 38 °C for 2 h.

After hydrolysis, 1.5 mL of ethyl acetate and 1.5 mL of n-hexane were added to each sample, mixed for 1 min using a vortex, and centrifuged at 1 °C for 10 min at 5000 rpm, for proper phase separation. The organic upper layer was then withdrawn, filtered, and transferred into vials for analysis.

The first standard solution was obtained by dissolving 10 mg of analyte in 100 mL of methyl alcohol, producing 100 mg/L of standard solutions. Analytical calibration curves were prepared by adding suitable amounts of more dilute standard solutions to “blank” urine samples. Specifically, 0.1 mL of 100 mg/L of standard solution was diluted to 10 mL with methyl alcohol, yielding 1.0 mg/L of standard solution. Subsequently, 5 mL urine samples were spiked with 5, 10, 25, 50, and 100 µL of 1.0 mg/L standard solution to achieve concentrations of 1, 2, 5, 10, and 20 µg/L. Additionally, each sample was spiked with 50 µL of BPA-D6, yielding 10 µg/L concentration. The blank urine was spiked only with BPA-D6 as ISTD. The calibration curves were developed with the ratio area BPs/area BPA-D6. An example of the chromatogram is shown in Figure 2. We optimized the mass analytical procedure by considering multiple transitions and selecting the most intense ones for method validation.

All results were normalized by the urine creatinine concentration, determined using a new HPLC-MS/MS procedure already validated in our laboratory [17].

3. Results

3.1. Performance of the Procedure

Retention times were as follows: BPS 5.75 min, BPF 7.3 min, and BPA 8.1 min.

The accuracy was assessed by analyzing a set of blank urine samples fortified with suitable amounts of pure standards to achieve exact nominal concentrations. It was determined as the mean percentage ratio of the found concentration, in µg/L, to the nominal concentration for each level, repeated three times in one day and in three non-consecutive days. Precision was evaluated using the coefficient of variation CV% of the triplicate QC analysis of a similar set of urine samples. Specifically, three QC samples at concentrations of 2 µg/L, 5 µg/L, and 20 µg/L were analyzed three times on the first day (Intraday Precision), and once again after three days and six days (interday precision). The validation data are shown in Table 1.

Precision was calculated as the repeatability of the analytical result, defined as the mean value of the CV% from five different sets of analyses, conducted on the same day (intraday), and on three different days (interday).

The linearity of the instrumental response with respect to the concentration levels was highly optimal. The corresponding R² values for the calibration curves were all 0.99 as shown in

Table 2. Calibration curves.

Analyte	Calibration Curve	R2	Range
BPA	y = 0.4453x + 0.0887	0.99	1–20 µg/L
BPS	y = 2.3018x + 0.9276	0.99	1–20 µg/L
BPF	y = 0.1197x − 0.0271	0.99	1–20 µg/L

.

3.2. Application to Real Samples

To evaluate the performance of the procedure, it was applied to a pool of male subjects attending a fertility center. These subjects were recruited as part of a more complex project aimed at evaluating possible exposure to endocrine disruptors, oxidative stress and infertility parameters. The Ethics Committee of IRCCS San Raffaele Institute in Milan approved the research protocol (approval number 18/INT/2023). From May 2023 to June 2024, 36 subjects were enrolled; each participant gave a written informed consent before participation. The characteristics of the analyzed population are presented in

Table 3. Characteristics of the analyzed male subjects (n = 36).

	n = 36
Age (range)	40.5 (32–52)
BMI a (% of subjects in the class)
Normal	44.4
Overweight	30.6
Obese	16.7
Unknown	8.3
Present smokers (%)	22.2
Previously smokers (%)	27.8
Alcohol consumption (%)
Daily	5.6
Weekly	13.9
Monthly	67.6
Never	12.0
Missing	4.6
Residence area (%)
Urban	91.7
Rural	8.3
Educational level
Lower middle school education	8.3
Upper secondary education	41.7
University degree or higher	50.0
Working activity
PC operators	41.7
Drivers	13.9
Craftsmen	2.8
Construction workers	5.6
Industrial workers	2.8
Health operators	11.1
Missing	22.1

^a BMI—body mass index.

.

The results shown in

Table 4. Results of the urine analysis of the recruited subjects.

	µg/L			µg/g Creatinine
	BPA 1	BPS 2	BPF 3	BPA 1	BPS 2	BPF 3
average	1.51	1.58	0.17	0.97	0.87	0.14
median	0.88	0.00	0.00	0.71	0.00	0.00
SD 4	2.28	6.80	0.42	1.27	3.03	0.34
minimum	0.00	0.00	0.00	0.00	0.00	0.00
maximum	10.29	40.80	1.79	6.23	17.13	1.25
% <LOD	25	66.67	75.00

¹ BPA—bisphenol A; ² BPS—bisphenol S; ³ BPF—bisphenol F; ⁴ SD—standard deviation.

, corrected for creatinine, revealed very clean chromatographic peaks, with very low noise and few interfering peaks. The quantification was quick and easy.

Most of the subjects showed relevant exposure to BPs, particularly to BPA. This is noteworthy since the use of this substance is subject to restrictions in Europe for numerous applications. It is therefore reasonable to conclude that this exposure is the result of the persistent, widespread diffusion of products containing BPA in the environment, with recycled plastics being a potential source.

All three BPs were found in the urine samples with variability between subjects.

An evaluation of possible differences between smokers, levels of education, work activities, and place of residence was conducted but did not show statistical significance (Mann–Whitney test for the residence and the Kruskal–Wallis for the other parameters).

To evaluate the potential health risks associated with BPA exposure in European residents, it is necessary to consider the biomonitoring guidance values (HBM-GV), particularly those relevant to urine samples. These values represent the maximum permissible levels to prevent adverse health effects. The reference value for urinary BPA in the general population, as defined in Germany, is 7 µg/L [18]. The results of the present study fall within this range, suggesting that the procedure presented here could be valuable for biomonitoring both exposed workers and the general population.

4. Discussion and Conclusions

A comparison between the analytical procedure presented here and those published in the literature for the analysis of BPs in human urine is presented in

Table 5. Comparison of validation parameters of different analytical methods for BP determination in human urine.

Reference	Analytes	Sample Preparation	Equipment	LOD 1,* (ng/mL)	Linear Concentration Range (ng/mL)	Mean Precision (CV 2 %)	Mean Accuracy (%)	Time of Analysis (min)
Our method	BPA 3, BPS 4, BPF 5	Enzymatic hydrolysis + 1 mL acetic acid; L/L 6 extraction with ethyl acetate and n-hexane.	LC-MS/MS 7	BPA = 0.01 BPS = 0.001 BPF = 0.07	BPA 0.03–20 BPS 0.003–20 BPF 0.20–20	BPA = 8.4 BPS = 7.8 BPF = 8.8	BPA = 109.0 BPS = 109.0 BPF = 97.8	9’
[17]	BPA, BPS, BPF	Enzymatic hydrolysis + 170 µL of water and 200 µL ammonium acetate.	LC-/MS/MS	BPA = 0.07 BPS = 0.07 BPF = 0.07	BPA 0.5–12.0 BPS 0.5–8.5 BPF 0.5–39.0	BPA = 10.8 BPS = 12.3 BPF = 9.8	BPA = 99.8 BPS = 94.2 BPF = 94.0	8
[18]	BPA, BPS, BPF, and others	Enzymatic hydrolysis + 1 mL ammonium acetate. L/L extraction with 1325 µL of CH3CN + 85 µL of tetrachloroethylene and 125 µL of acetic anhydride.	GC 8-MS	BPA = 0.03 BPS = 5 BPF = 0.03	BPA 0.1–50 BPS 2.5–50 BPF 0.5–50	BPA = 14.0 BPS = 5.0 BPF = 9.8	BPA = 87.0 BPS = 94.2 BPF = 73.0	13
[7]	BPA, BPS, BPF, and others	Enzymatic hydrolysis + dilution to 10 mL with aqueous solution of NaCl 20% and air assisted liquid/liquid micro extraction with 750 µL of 1,2 dichloroethane.	LC-MS/MS	BPA = 0.03 BPS = 0.02 BPF = 0.08	BPA 0.1–20 BPS 0.07–20 BPF 0.25–20	BPA = 7.2 BPS = 5.7 BPF = 3.4	BPA = 100.2 BPS = 101.7 BPF = 100.2	10
[19]	BPA, BPS, BPF, and others	Enzymatic hydrolysis + dilution to 465 µL with water. Addition of formic acid (400 µL, 0.5% solution); the sample was filtered using first a 0.45 µm and after a 0.2 µm filters. SPE 9 extraction.	LC-MS/MS	BPA = 0.1 BPS = 0.07 BPF = 0.4	BPA 0.1–200 BPS 0.5–200 BPF 0.5–200	BPA = 11.3 BPS = 16.1 BPF = 16.9	BPA = 103.0 BPS = 106.0 BPF = 103.5	8
[20]	BPA, BPS, BPF, and others	Enzymatic hydrolysis + 1 mL sodium acetate buffer (0.2 M). Then, L/L extraction with 2 mL of CH3CN and 0.3 mL ethyl acetate. Then the supernatant was evaporated to dryness under nitrogen stream and reconstituted with 500 µL methanol/water (50:50 v/v).	LC-MS/MS	BPA = 0.09 BPS = 0.01 BPF = 0.1	BPA 0.27–6.25 BPS 0.032–1.25 BPF 0.31–12.5	less than 16.4% for each BPs	BPA = 100.2 BPS = 93.4 BPF = 93.4	8
[4]	BPA and others	Enzymatic hydrolysis + 300 µL ammonium acetate (1 M). Frozen overnight, centrifuged, and the supernatant was analyzed.	LC-MS/MS	BPA = 0.2	BPA = 0.5–80.0	BPA = 4.5	BPA = 108.3	16.1
[12]	BPA, BPS, and others	Enzymatic hydrolysis + 200 µL ammonium acetate. L/L extraction with methyl tert-butyl ether/ethyl acetate (5:1, v/v) 3 times. Then, the supernatant was evaporated to dryness under nitrogen stream, then reconstituted with 200 µL acetonitrile/water (60:40 v/v), then centrifuged.	UHPLC 10-MS/MS	BPA = 0.2 BPS = 0.2	BPA 1.00–50.00 BPS 1.00–50.00	BPA = 8.9 BPS = 4.2	BPA = 95.2 BPS = 101.1	8
[13]	BPA, BPS, and others	Enzymatic hydrolysis + 1.5 mL ammonium acetate. Then, 10 mL of 7% NaCl aqueous solution (w/v) at pH = 2 with HCl 0.5 M was added. Dispersive L/L micro extraction with 0.5 mL of acetone and 0.75 mL of trichloromethane. The organic phase was evaporated under nitrogen stream. The residue was dissolved with 100 µL of 60:40 methanol (0.1%ammonia)/water (0.1%ammonia).	UHPLC-MS/MS	BPA = 0.2 BPS = 0.1	BPA = 0.5–40.0 BPS = 0.5–40.0	less than 14% for each BPs	BPA = 99.3 BPS = 100.7	5
[14]	BPA, BPS, BPF	Enzymatic hydrolysis + 80 µL 1 M of formic acid and 670 µL of water were added to 100 µL of the mixture and applied to an online SPE.	LC-MS/MS	BPA = 0.2 BPS = 0.1 BPF = 0.2	BPA 1–50 BPS 1–50 BPF 1–50	BPA = 4.8 BPS = 6.6 BPF = 4.8	BPA = 105.1 BPS = 102.9 BPF = 106.1	12
[21]	BPA	Enzymatic hydrolysis + 5 mL of buffer solution (citric acid and sodium hydroxide). SPE.	UPLC-MS/MS	BPA = 0.1	BPA = 0.05–25	BPA = 2.2	BPA = 79.0	6

* LODs were rounded to the first significant digit. ¹ LOD—limit of detection; ² CV—coefficient of variation; ³ BPA—bisphenol A; ⁴ BPS—bisphenol S; ⁵ BPF—bisphenol F; ⁶ L/L liquid/liquid; ⁷ LC MS/MS—liquid chromatography–tandem mass spectrometry; ⁸ GC—gas chromatography; ⁹ SPE—solid phase extraction; ¹⁰ UHPLC—ultra high-pressure liquid chromatography.

. The comparison shows that the pre-treatment stage of the samples varies significantly across different methods.

Some authors have pointed out [7] that potential stages of sample contamination, due to the release of BPs from laboratory plastics, can occur at various points during sample treatment—from collection to solid phase extraction—and even during the chromatographic run.

For this reason, proposing a phase of pre-treatment of the sample that is simple, involves fewer steps, and minimizes the contact of samples with plastic materials, seems to be a suitable choice.

The procedure herein presented involves, after enzymatic hydrolysis, exclusively a liquid/liquid extraction, conducted using laboratory glassware, which effectively reduces the risk of contamination.

As shown in Table 5, the choice of multiple sample pre-treatment steps does not necessarily produce better results in terms of method performance.

The first aspect to be highlighted is that the LODs of our procedure are significantly lower than those reported in the literature.

In particular, for BPA, our procedure achieves LODs that are 3 [20] to 20 [15,22,23] times lower than those found in other studies; for BPS, the LOD is 10 to 200 times lower (without considering the only work in gas chromatography [20] whose LOD is higher); for BPF, the value is comparable to some studies [12,19,20] but up to 5 times lower than others [14].

This evidence confirms the effectiveness of the sample pre-treatment procedure, as it allows for a significant reduction in background noise.

It is possible that the particular care taken to avoid possible background contamination during sample handling and analysis contributed to obtaining better LOD and LOQ values, which are linked to a reduction in background noise.

When examining the linearity ranges, the procedure presented here shows values comparable with other published studies [7,13,19]. However, it should be noted that the lower limit of the curve is lower than in other papers, due to the previously discussed lower LOD values. In some studies, the upper limit of the calibration curve is higher [6,14,15,20,21,22] than that proposed here, with one study even reporting an upper limit an order of magnitude higher [14]. However, it is crucial to consider the specific purpose for which the method is intended when evaluating these differences.

The new, presented analytical procedure is intended for application in biological monitoring of both the general population and potentially exposed workers. In these cases, the expected concentrations fall within our linear dynamic range. In fact, the proposed reference value for the general population is of a few units, for example, for the BPA of 7 µg/L [18] in Germany.

Considering that BPs may produce adverse effects even at low doses [15], having an analytical procedure capable of accurately measuring low concentrations is particularly advantageous.

Comparing the mean analytical precision values, some studies have reported better values for all the BPs [13,23] but in most cases, the average precision values are worse than ours [13,14,19,22]. In some cases, the precision of the new procedure is better for BPA [6,7] or for BPS but not for BPA and BPF [12]. This indicates that the precision associated with the new procedure is largely comparable with the literature and, in some cases, even better.

Regarding the accuracy, the results fall within the acceptability range ±20% for all BPs. For BPA, the accuracy is largely comparable to the other published values. For BPF, it is superior to that reported in other studies [12,14,19,21,23] while, for BPS, although the deviation from 100% is higher than in other methods, this difference is often only a few percentage points.

BPS responds more sensitively in mass spectrometry compared to the other two alkylphenols, resulting in larger peaks which may contribute to a slightly greater deviation from ideal accuracy.

One limitation, for the interpretation of data but not for the method performance, is the use of a single spot urine sample. While this is an efficient sampling method, it could introduce bias due to variations in physiological status, daily activities, and specific exposures among individuals. In terms of biomonitoring of the population, it should be underlined that the application of the method in the present study involved an exclusively male population; certainly the preparation of investigations also on the female population is desirable for a more specific exposure evaluation, particularly if oriented towards the evaluation of health effects.

The main disadvantage of this analytical procedure could be the intrinsic volatility of the solvents used in the extraction procedure, specifically n-hexane and ethyl acetate, that could affect the resulting extraction volume. To mitigate this issue, it is strongly recommended to perform both the extraction and centrifugation phases at low temperatures. However, the use of a deuterated internal standard can compensate for any potential variations, ensuring the reliability of the final analytical results.

Furthermore, the completion of the evaluation of the validity of the use of the proposed method will require the involvement of other laboratories, to further evaluate the robustness of the procedure.

In conclusion, this study presents a fast, simple, sensitive, and selective analytical procedure for determining the most produced bisphenols in industry.

Among the benefits of this analytical procedure are its simplicity and the speed of the overall process, with the enzyme hydrolysis phase being the only time-consuming step. Additionally, the method’s high sensitivity, as demonstrated by its low LODs and LOQs, is noteworthy.

The analytical procedure offers several advantages, foremost being the lower LODs for all three substances, permitting the quantification of lower levels than previously possible. This is particularly important given the ability of these BPs to act as endocrine disrupters (either confirmed or suspected) with the possibility of producing adverse health effects even at low concentrations. The procedure is well suited for biological monitoring and screening in both the general population and exposed workers.