Twelve volunteers were recruited at the University of Freiburg, Germany to take part in the study. Inclusion criteria was an age of at least 20 years and a good physical constitution. Exclusion criteria for participation were the presence of pregnancy or serious illnesses such as cardiovascular, respiratory, metabolic, liver or kidney disease, mental or neurological illnesses or addiction. The presence of diseases was recorded by questionnaire at the beginning of the study. Serious illness led directly to exclusion from the study. To screen for alcohol abuse, the average alcohol consumption per week was surveyed and the indirect alcohol markers GGT and MCV were determined. Also a single analyzes of the direct alcohol marker PEth was also conducted. Participation was voluntary and the participation in the study could be cancelled at any time. Table 1 contains data of the participants.

The drinking study took place at the Institute of Forensic Medicine at the University Medical Centre Freiburg from April to July 2023. Before the start of the trial, all volunteers were interviewed by a doctor and informed about possible risks. An initial sample collection was done (as explained below). All participants underwent a four-week abstinence phase, followed by two drinking trials (trial A and B) with targeted blood alcohol concentrations (BAC) of 0.6 g/kg and 0.75 g/kg and subsequent three-day sample collection in a private setting. There was a minimum period of nine days between the two drinking trials. The time schedule of the trial is shown schematically in Fig. 1. The individually required amount of alcohol (Wodka Gorbatschow, 37.5vol%, mixed with fruit juice) was calculated using the Widmark formula and was consumed within 30 min (1st drinking event) or 45 min (2nd drinking event). The participants’ last meal was consumed no later than two hours before the start of the drinking event. Once the target BAC was reached, the participants were kept under medical supervision at the Institute of Forensic Medicine until a breath alcohol concentration of less than 0.05 mg/L was reached.

Venous blood samples were collected at the beginning of the abstinence phase, to identify (and potentially exclude) participants with alcohol abuse by analyzing MCV and GGT as well as a single determination of a baseline value for PEth as a dried-blood spot sample (DBS) and, if necessary, follow-up checks after 3 weeks. The determination of PEth from capillary blood of the fingertip and dried on a filter paper card as DBS was carried out due to the simplified storage and sample collection. The result accuracy of DBS is comparable to venous blood samples [5, 25, 26]. DBS sampling is easier to handle during the trial, as it can be done by self-sampling by the volunteers “at home” on the days after the supervised drinking day. DBS samples were collected on DBSV cards (Greencheck DBSV, Protzek, Lörrach, Germany). In addition, before the start of both drinking trials, urine samples were collected for determination of EtG by LC-MS/MS with a limit of detection (LOD) of 100 ng/mL to confirm abstinence prior to the experiment. Women took a pregnancy test (β-HCG) before consuming alcohol. In each of the two drinking trials, one urine sample was taken to determine EtG, seven capillary blood samples were taken as DBS to determine PEth and one whole blood sample was taken to determine the BAC 45 min after the end of drinking. On the day of the drinking event (later referred to as “day zero”), a urine sample (U NP) and a DBS sample (DBS NP) were taken before the start of the drinking test. Breath alcohol was monitored at 15-minute intervals from 45 min after the end of drinking on throughout the study day until it fell below an AAC of 0.05 g/l. A venous whole blood sample (vBAC) was obtained 45 min after the end of drinking to determine the BAC, from this sample a DBS (DBS vBAC) was prepared. Further DBS from capillary blood were taken 30 min after the breath alcohol concentration fell below 0.05 g/l (DBS 0) and in the evening at around 9 pm (DBS 1) by self-sampling in a private setting without supervision. In addition, four capillary blood DBS samples were taken by the participants on the three following days, one in the morning and one in the evening on day 1 (DBS 2–3) and one in the morning on day 2 (DBS 4) and on day 3 (DBS 5). Figure 1 provides an overview of the sample collection. The schedule described in detail for the first trial was also applied for the second trial.

Time schedule and sample collection. U: Urine sample, NP: Sample before start of drinking, DBS: Dried blood spot, vBAC: venous blood alcohol concentration

All solvents used were of HPLC quality. Methanol (MeOH) (≥ 99.9%) was obtained from Biosolve BV (Valkenswaard, Netherlands), acetonitrile (MeCN) (99.9%) from Acros Organics (New Jersey, USA) and isopropanol (≥ 99.5%) from Fisher Scientific (Loughborough, UK). Ammonium acetate (Fractopur) was purchased from Merck (Darmstadt, Germany). Sigma-Aldrich (Buchs, Switzerland) provided formic acid (FA) (50%).

PEth 16:0/18:1 was analyzed on DBS by LC-MS/MS as described elsewhere [26, 27]. This type of DBS card system was earlier compared to classical DBS cards [28]. In brief, for each sample, one paper tooth was detached from its holder and placed in a 2 mL plastic vial (Sarsted, Nümbrecht, Germany). MeOH (1 mL) and pentadeuterated internal standard solution (10 µL) were added. The samples were then shaken for 20 min (IKA Vibrax, Staufen, Germany) and subsequently centrifuged for 10 min at 13,000 rpm and 8 °C (Mikro 220, Hettich, Tuttlingen, Germany). Then, the extract was transferred to 2 mL auto sampler glass vials (Wicom, Heppenheim, Germany) and evaporated at 50 °C under a gentle stream of N2. After reconstitution with MeOH (100 µL), the sample (injection volume 4.5 µL) was analyzed using an LC-MS/MS system consisting of an UltiMate® 3000 HPLC system (Dionex, Thermo Scientific Instruments, Reinach, Switzerland) coupled to a 5500 QTRAP with a TurboIonSpray source (Sciex, Toronto, Canada) operated in negative ionization mode using a previously published method [26]. The limit of quantitation (LOQ) was 10 ng/mL, LOD was set at 5 ng/mL, and calibration ranged from 7.5 to 1500 ng/mL.

EtG as well as EtS in urine were analyzed by LC-MS/MS as described by Luginbühl et al. [29]. The limit of quantitation was at 0.1 mg/L. In short, 150 µL of working solution (300 µL EtG-D5 (0.1 mg/mL, Cerilliant, Round Rock, Texas, USA) and 50 µL EtS-D5 (0.1 mg/mL, Lipomed, Arlesheim, Switzerland) in 25 mL MeCN) was added to 50 µL of urine sample, which was then subsequently shaken for 5 min and centrifuged for 10 min at 8 °C and 13,000 rpm. 100 µL of supernatant were transferred to an auto sampler glass vial and evaporated at 50 °C under a stream of N2. After reconstitution in 100 µl water/ MeCN (95/5) with 0.1% FA, the sample was analyzed using an LC-MS/MS system consisting of an UltiMate ® 3000 HPLC system (Dionex, Thermo Scientific Instruments, Reinach, Switzerland) coupled to a 3200 QTRAP instrument (Sciex, Toronto, Canada) operated in negative ionization mode using a previously published method [29]. LOQ was 0.1 mg/L, and calibration ranged from 0.1 to 10 mg/L.

The determination of the blood alcohol concentration (BAC) from serum was carried out at the Institute of Forensic Medicine at the Freiburg University Medical Centre using headspace gas chromatography with flame ionization detection (HS-GC-FID). After centrifugation of the blood samples, 100 µL supernatant with 500 µL tert-butanol solution (internal standard) was pipetted into a 20 mL headspace vial. The analysis was performed on two GC-FID systems: A Clarus 580 with an Rtx®-BAC Plus 1 column and a Clarus 680 with an Rtx®-502.2 column, both coupled to TurboMatrix headspace samplers (Perkin Elmer, Rodgau, Germany).