Seventeen healthy adults (nine men and eight women) voluntarily participated in the study. Inclusion criteria required participants to be of either gender, aged between 18 and 45 years, without significant medical conditions or ongoing treatments that could interfere with the study drug’s safety or pharmacokinetics, with no history of psychiatric disorders, and with previous experience with recreational intranasal use of psychostimulants and/or synthetic cathinones. Subjects older than 45 years were excluded because this group of substances are used between 18 and 40 years and to avoid the possible progressive natural physiological and pharmacokinetic changes related to aging and to increase safety. Exclusion criteria included a history of significant medical or psychiatric disorders (with the exception of nicotine use disorder), prior severe adverse reactions to psychostimulants, current long-term medication use, and pregnancy.

Participants were recruited through word of mouth in collaboration with Energy Control (Asociación Bienestar y Desarrollo, https://energycontrol.org/), a harm-reduction non-government organization (NGO) that provides counseling and drug-checking services. The study protocol was approved by the local Human Research Ethics Committee (CEIC HUGTiP, Badalona, Spain; ref. PI-18–267). All participants received detailed information about the study’s objectives and procedures and provided written informed consent prior to any study-related activity. The study was conducted in accordance with the Declaration of Helsinki and applicable Spanish research regulations (Biomedical Research Law 14/2007). Participants were financially compensated for their time and involvement.

We conducted a prospective, non-controlled observational study in a naturalistic setting. The study design, including procedures and assessment methods, was consistent with those applied in our previous research on the acute effects of other NPS (De la Rosa et al., 2025; Poyatos et al., 2021; Papaseit et al., 2021; Martínez et al., 2021; Papaseit et al., 2020a). Participants independently acquired the substance from unidentified suppliers, and the products were analysed by Energy Control. Participants were allowed to select their own doses within predefined ranges based on common use patterns reported in the literature: 75–90 mg for cocaine and 15–25 mg for α-PHiP. The selected doses were determined according to participants’ prior consumption experience. Specifically, five male participants selected 90 mg of cocaine, while three female participants chose 75 mg. For α-PHiP, five females and one male selected 15 mg, whereas three males opted for 25 mg.

Participants were asked to abstain from consuming any recreational substances during the week prior to the selection visit, and to avoid alcohol, energy drinks and caffeinated beverages within 24 h before the experimental session, in order to minimize the risk of interactions or residual effects. After a selection visit, which was used to confirm eligibility, two experimental sessions (one for each substance) were conducted. During the selection visit, participants were asked to provide information on their medical history (including mental health disorders), drug consumption history, and underwent a physical examination and a urine drug test. The selection visit took place 2–3 days before the session.

On the day of the session, participants were asked to arrive at 15:00 at a private club in Barcelona, with the aim to simulate a recreational consumption environment. The study setting was closed to the public. Each session had a duration of 5 h, beginning immediately after the self-administration of the selected substance.

Upon arrival, participants were required to provide a urine for a drug screening test to confirm abstinence from prior substance use. The screening tested for the presence of benzodiazepines, barbiturates, morphine, methadone, cocaine, amphetamines, methamphetamine, MDMA, THC, and tricyclic antidepressants using the Drug-Screen Multi 10TD Test (Multi-Line, Nal Von Minden, Moers, Germany). Previous alcohol intake was ruled out by assessing breath alcohol concentration (Drager Alcotest 5820, Dragerwerk AG & Co., Lübeck, Germany). In addition, Female participants underwent a urine pregnancy test (Glip Test Plus hCG Card®, Ref 30,701, Biosynex, Delémont, Switzerland), which was required to be negative for inclusion in the study.

During the session, participants were allowed to engage in leisure activities of their choice, such as listening to music, dancing, or conversing. Assessments were performed at baseline (prior to self-administration) and along 5 h after intranasal self-administration of α-PHiP or cocaine. A light snack (a piece of fruit) was provided 2 h after administration.

At each time point, assessments followed a predefined sequence: i) oral fluid (OF) sample collection (from 5 min to the scheduled time point), ii) measurement of vital signs (from 5 min to the scheduled time point) and iii) completion of questionnaires (from the scheduled time point until 5 min later). Capillary blood at 1 h and urine samples at 2 h and 5 h following administration were collected after the completion of the questionnaires.

In both sessions OF samples were collected using Salivette tubes (Sarstedt AG & Co. KG, Nümbrecht, Germany) before and following substance self-consumption at 0.33 h (20 min) and 0.66 h (40 min), 1, 1.5, 2, 2.5, 3, 4, and 5 h. Samples were centrifuged after collection and stored at −20 °C until analysis. Urine was collected in two different time ranges: 0–2 and 2–5 h following consumption. The total volume was registered, and a sample was collected in 1.5-mL tubes and stored at −20 °C prior to analysis. Sweat collection patches (Cosmopor E, Hartmann, Heidenheim, Germany) were applied to the backs of the subjects, under the scapula, 4 h before administration (−4 to 0 h) and up to 5 h after self-administration (0–5 h). Dry blood samples were collected via finger prick (Whatman filter paper, Maidstone, UK), where a blood drop filling a pre-marked circle corresponded to approximately 30 µL of blood. DBS samples were collected 1 h after consumption and stored at ambient temperature until analysis.

Noninvasive heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP) and cutaneous temperature were measured using an automatic monitor (Omron, Hoofdorp, Netherlands) with subjects seated at baseline, and at 20 min, 40 min, 1 h, 1.5 h, 2 h, 2.5 h, 3 h, 4 h, and 5 h after administration.

Subjective effects were assessed at baseline, 0.33 h (20 min), 0.66 h (40 min), 1 h, 1.5 h, 2 h, 2.5 h, 3 h, 4 h, and 5 h after administration, using a series of visual analog scales (VASs) and the Addiction Research Center Inventory_ARCI 49 item-short form (same time points except 20 min). The Evaluation of Subjective Effects of Substances with Abuse Potential questionnaire (VESSPA-SSE) was administered at baseline, 1, 2, 3, 4, and 5 h. The different tools were administered in the following order (VASs, ARCI and VESSPA). Additionally, the Positive and Negative Syndrome Scale for Schizophrenia (PANSS) was administered at baseline and at 5 h (De la Rosa et al., 2025; Poyatos et al., 2023).

α-PHiP and α-Pyrrolidinovalerophenone-d8 (α-PVP-d8) as internal standard (IS) were purchased from Cayman Chemicals (Ann Arbor, MI, USA). LC-MS grade water and analytical grade ethyl acetate and other chemicals (sodium phosphate and sodium hydrogen phosphate, sodium carbonate and sodium hydrogen carbonate) were obtained from Carlo Erba (Cornaredo, Italy).

Working standard solutions of α-PHiP (10, 1 and 0.1 μg/mL) and the IS (10 and 1 μg/mL) solution were prepared by appropriate methanolic dilution of stock solution. OF calibrators were prepared by adding appropriate volumes of the working standard solutions to blank OF to cover a calibration range of 5–1,000 ng/mL. Blank urine and DBS samples were spiked in the range of 5–500 ng/mL, while blank sweat patches was spiked in the range of 5–500 ng/patch.

Low- and medium-quality control (QC) samples were prepared at concentrations of 7.5 and 250 ng/mL or ng/patch, respectively, OF, urine, DBS and sweat patches samples. High-QC samples were prepared at 850 ng/mL for OF, and at 450 ng/mL or ng/patch, for urine, DBS, and sweat patches samples, respectively.

A similar extraction procedure was carried out for all matrices. Aliquots of (50 µL), urine (100 µL), DBS (30 µL), and sweat (with the absorbent pad removed from the patch using clean tweezers) were transferred into clean glass tubes and spiked with the appropriate volume of IS working solution. Specifically, 5, 10, and 3 µL of a 1 μg/mL IS solution were added to saliva, urine, and DBS samples, respectively, while 10 µL of a 10 μg/mL IS solution were added to sweat samples, to achieve a final concentration of 100 ng/mL or ng/patch for saliva, urine, DBS, and sweat, respectively.

After the addition of 200 µL (for OF and urine) or 1.5 mL (for sweat and DBS) of a carbonate/bicarbonate buffer solution (0.8 M, pH 9.0), sweat and DBS samples were sonicated for 10 min prior to liquid–liquid extraction (LLE) with 1 mL of ethyl acetate using a horizontal shaker for 10 min. OF and urine samples were subjected directly to LLE. Following extraction, the samples were centrifuged at 4,000 rpm for 5 min. The organic phase was collected and evaporated to dryness under a gentle stream of nitrogen. The resulting residues were reconstituted in 50 µL of ethyl acetate, transferred to autosampler vials, and 0.5 µL was injected into the GC-MS/MS system for α-PHiP quantification.

Analysis was performed on a 7890B GC system equipped with a multimode inlet (MMI) and a 7693A Automatic Sampler (all from Agilent Technologies, Santa Clara, CA, USA). The capillary column used was an Agilent Technologies HP-5MS UI (30 m × 0.25 mm × 0.25 m). The samples were injected in pulsed splitless mode with the injector port temperature of 260 °C. Helium was used as a quenching gas at a flow of 2.25 mL/min The column temperature was initially set at 80 °C for 1 min before increasing to 160 °C at 20 °C/min, to 250 °C at 10 °C/min and then increasing again to 290 °C at 35 °C/min (held for 1 min), taking the total run time to 16.14 min.

The GC instrument was interfaced to a 7000D GC/MS triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA). The MS/MS analyses were conducted in positive electron ionization (EI) mode. The transfer line and ion source temperature were set at 280  C. Nitrogen was used as the collision-induced dissociation (CID) gas for ion fragmentation at a flow of 1.5 mL/min.

Quantifications were performed using multiple reaction monitoring (MRM) transitions. Two transitions for each analyte and one transition for the deuterated standards were selected. The α-PHiP quantifier transition was m/z 140 > 98 (Collision energy: 14 eV) and its qualifier transition was 140 > 84 m/z (Collision energy: 10 eV), while α-PVP-d8 transition was 134 > 105 m/z (Collision energy: 14 eV).

Prior to application to real samples, the analytical method was assessed for linearity, sensitivity, accuracy, precision, carryover, and dilution integrity in accordance with the guidelines established by the Organization of Scientific Area Committees for Forensic Science (OSAC, 2019). The validation protocol encompassed the evaluation of bias, precision (within-run and between-run), calibration model, limit of detection (LOD), lower limit of quantification (LLOQ), carryover, matrix interferences, and dilution integrity, as previously described (Di Trana A et al., 2025).

In brief, linearity was assessed using six calibration curves, each comprising six points, over five different days. The acceptance criteria required quantification within ±15% of the target, a coefficient of determination (R ²) ≥ 0.99, and a quantifier/qualifier ratio within ±20%. Additionally, the Mandel test was performed to confirm linearity. Sensitivity was evaluated by determining the LOD and LLOQ. The LOD was defined as the lowest concentration showing a peak within ±0.1 min of the average retention time and a signal-to-noise ratio ≥3, using diluted fortified blank matrices. The LLOQ was established at the lowest non-zero calibrator, with retention time and quantification within ±0.1 min and ±20%, respectively.

Bias was calculated as the percent deviation of the mean measured concentration from the nominal value across five independent runs at three concentration levels (low, medium, and high QC). An acceptable bias was within ±20%. Precision was expressed as the coefficient of variation (%CV) and evaluated by triplicate analysis of QC samples across five runs within the same day and over 5 days. A %CV <20% was considered acceptable. Carryover was assessed by injecting blank matrix samples immediately following the highest calibrator. The method was deemed free of carryover if no analyte signal with a signal-to-noise ratio ≥3 appeared within ±0.1 min of the calibrator retention time.

Dilution integrity was verified by analysing samples spiked above the upper calibration limit (up to 20-fold) and diluted to fall within the validated range. A % CV <20% was considered acceptable.

The quantitative methods for α-PHiP were successfully validated (see Supplementary Table 1). Each assessed parameter proved to be within the acceptable criteria proposed by OSAC for Forensic Sciences guidelines (OSAC, 2019).

Aliquots OF (50 µL), urine (50 µL), and sweat (with the absorbent pad removed from the patch using clean tweezers) were transferred into clean glass tubes and spiked with the appropriate volume of IS working solution to achieve a concentration of 100 ng/mL or ng/patch for OF, urine, and sweat, respectively.

Following the addition of 500 µL of phosphate buffer pH 6.0 for OF and urine, or 2 mL for sweat, all samples underwent solid-phase extraction (SPE) using Strata™-X-C cation exchange cartridges (Phenomenex, Torrance, CA, USA). Cartridges were conditioned with 2 mL of methanol, 2 mL of water, and 2 mL of phosphate buffer (pH 6.0). Samples were then loaded, washed sequentially with 2 mL of water and 1 mL of 0.1 M HCl, followed by 2 mL of methanol, and finally eluted with 2 mL of a dichloromethane/isopropanol/ammonium hydroxide mixture (77:20:3, v/v/v). The eluates were evaporated to dryness and derivatized with 50 µL of N, O Bis(trimethylsilyl)trifluoroacetamide solution with 1% of trimethylchlorosilane (BSTFA+ 1%TMS) in a heating block at 70 °C for 30 min prior to GC-MS/MS analysis. 1 μL was injected to Cocaine (COC) and its metabolite benzoylecgonine (BZE) determination.

COC and BZE were detected using a previously published and validated gas chromatography–tandem mass spectrometry (GC-MS/MS) method (La Maida et al., 2021). The analytes were separated by gas chromatography and detected using a triple quadrupole mass spectrometer operating in multiple reaction monitoring (MRM) mode with positive electron ionization. MRM transitions were 182→82 (collision energy 10 eV) and 182→108 (collision energy 10 eV) for COC, and 361→82 (collision energy 20 eV) and 240→82 (collision energy 20 eV) for BZE. Underline transition was used for the quantification.

The method was validated (see Supplementary Table 2) according to international guidelines and demonstrated acceptable performance in terms of linearity (R ² ≥ 0.990). The limits of quantification (LOQ) were 10 ng/mL or ng/patch and 1 ng/mL or ng/patch for COC and BZE, respectively and adequate for the purpose of the present study. Intra- and inter-assay precision were always below 20.0%, and accuracy deviations did not exceed ±10.3%. Dilution integrity was assessed, and over-curve sample concentrations remained within ±15% of the target values for all analytes.

In the OF samples, cortisol levels were also quantified to explore potential stress-related or metabolic interactions with α-PHiP and cocaine exposure. Salivary cortisol was quantified using electrochemiluminescence immunoassay (ECLIA) method (Elecsys Cortisol II), performed on the Cobas e analyzer platform (Roche Diagnostics GmbH, Mannheim, Germany).

The determination of the sample size was based on the methodology of bioequivalence studies, which resulted in eight to nine subjects per substance needed, considering an alpha risk of 0.05, a power of 80%, and a difference of at least 35% between 20 min and baseline values in the intensity/high effect, considering a 25% of variability.

Peak concentration (C_max) and time taken to reach peak concentration (T_max) in OF were directly obtained from the concentration–time curves. The area under the curve (AUC_0–5h) was calculated by the linear trapezoidal rule. The terminal-phase elimination half-life (t_1/2) was calculated as 0.693/Ke, where Ke (elimination rate constant) was the slope of the apparent elimination phase of the natural logarithmic (ln) transformation of the OF concentration–time curve, which was estimated using linear regression.

Differences with respect to baseline were calculated for vital signs and subjective effects. Maximum effects (E_max) and the time needed to reach maximum effects (T_max) were also calculated for these outcomes. The areas under the curve of the effects (AUC_0–5h) were calculated also using the trapezoidal rule.

Two-way analysis of variance (ANOVA) was conducted to evaluate the influence of dose and gender/sex on the OF concentrations and effects (E_max and AUC values). A gender/sex effect was observed in several PHiP subjective outcomes (32% of them had higher values in men who received a higher dose). Due to our limited sample size, we decided to include all the participants in one group to compare with cocaine, and we provide the gender differences in a supplementary table (see Supplementary Table 3).

The comparison between the effects of α-PHiP and cocaine was performed with an independent sample T-test (E_max and AUC). We also conducted a Dunnett post hoc test to compare the different time points with baseline values, which was adjusted for multiple comparisons. T_max values were compared with the non-parametric Mann-Whitney U-test.

Statistical analysis was carried out using PASW Statistics version 18 (SPSS Inc., Chicago, IL, United States). Differences were considered statistically significant when the resulting p value was <0.05.

A total of 17 subjects (nine males and eight females) participated in the study. Their mean age was 31.82 ± 5.96 years (range: 24–46 years), mean weighted 66.04 ± 12.78 kg (range: 44–80 kg), and mean body mass index (BMI) was 22.46 ± 3.47 kg/m² (range: 20–30 kg/m²).

All selected participants reported prior experience with psychostimulants, including cathinones, cocaine, MDMA, amphetamines, methamphetamines and 4-bromo-2,5- dimethoxyphenethylamine (2-CB). They reported a median of 218 (range 27–1,156) lifetime uses, a median of 51 (range 3–255) uses in the past year and a median of 4 (range 0–18) uses in the past month. Six were current tobacco smokers, and all reported alcohol consumption. At the beginning of the experimental session, all participants tested negative on urine drug screening, and no clinical signs of intoxication were observed at baseline.

The summary of the parameters of physiological outcomes can be seen in Table 1. Furthermore, it includes statistically significant comparisons to baseline using the Dunnett test for the different time assessments. Additionally, Figure 1 shows the time course of heart rate and systolic blood pressure.

Regarding vital signs, both substances produced a significant increase in heart rate compared to baseline, with a trend for a higher increase with the synthetic cathinone. α-PHiP induced a sustained and statistically significant elevation of SBP over the 5 h, whereas for cocaine it was observed during the first hour. A comparable pattern was observed for diastolic blood pressure, showing also similar peak values among both substances.

Finally, no significant effect on body temperature was detected after α-PHiP administration, while cocaine produced a sustained increase across the 5-h monitoring period.

Table 2 presents the pharmacodynamic parameters for all the subjective outcomes. α-PHiP and cocaine produced psychostimulant subjective effects collected through VAS, ARCI and VESSPA-SSE. Overall, subjects reported subjective effects starting at 20 min with maximum values within the first hour. Subjective effects of both substances disappeared from 3 to 5 h. Although the maximum effects were similar, cocaine produced greater intensity and more pronounced high effects than α-PHiP. The synthetic cathinone mainly produced good effects, liking, clarity, focused, open to others, trust to others and would like to be with other people.

In the ARCI questionnaire, significant differences from baseline were found in the MBG (morphine benzedrine group, euphoria), BG (benzedrine group, intellectual efficiency and energy), and A (amphetamine-like effects) subscales, in both substances. Figure 2 shows a selection of the main subjective effects reported with both substances.

Regarding the VESSPA-SSE, α-PHiP caused significant changes compared to baseline in several subscales, such as ANX (anxiety), SOC (pleasure and sociability), ACT (activity and energy), in similar proportion than with cocaine.

No serious adverse effects were reported. No changes were observed in the PANNS score during the sessions. Four participants reported an unpleasant sensation in the throat/numbness after snorting cocaine, which can be explained by its local anaesthetic effect.

Across all participants, α-PHiP was rapidly absorbed, and the time to reach maximum concentrations was shorter than for cocaine, 0.33 vs. 0.55 h, respectively (see Table 3). Mean concentration time curve for α-PHiP in OF is shown in Figure 3. After 5 h, α-PHiP concentrations were near-baseline levels. α-PHiP had median salivary half-life of 1.06 h (0.84–1.79, with an oulier of 5.83) while for cocaine it was 0.84 h and for BZE 3.22 h. Cocaine concentrations in OF were more than 5-fold higher than those of α-PHiP, while BZE concentrations were significantly lower in this matrix. Two volunteers had very high cocaine concentrations (>5,000 ng/mL) within the first hour suggesting contamination during the administration.

Although a slight increase in salivary cortisol was observed following administration, this change did not reach statistical significance across the study population, indicating that disposition in oral fluid occurred independently of measurable alterations in cortisol secretion dynamics.

The total amount of α-PHiP excreted in urine over the 5-h period varied considerably among individuals, ranging from approximately 0.004–0.06 mg and only about 0.11% of the intranasally administered dose was excreted unchanged in urine. In detail, a total α-PHiP mean amount of 21.6 ± 16.8 µg was accumulated in urine up to 5 h, of which 6.4 ± 7.9 µg and 15.3 ± 16.1 µg between 0–2 and 2–5 h, respectively.

Almost all volunteers exhibited very low or undetectable levels of α-PHiP in sweat samples collected over the 0–5 h interval. Only three participants demonstrated appreciable drug concentrations within this timeframe. One individual presented an exceptionally high value of 444.2 ng/patch, another exhibited a notably elevated level of 76.6 ng/patch, and a third showed a moderate concentration of 30.0 ng/patch. On average, approximately 0.09 ± 0.20 mg of α-PHiP was excreted in sweat after administration, equivalent to around 0.6% of the total intranasally administered dose.

α-PHiP levels in DBS samples collected 1 h after intranasal administration varied notably across volunteers. Most volunteers exhibited concentrations ranging from 8.3 to 24.1 ng/mL. However, one volunteer showed a markedly elevated concentration of 446.9 ng/mL, substantially exceeding values observed in all other participants.