Detection of Anabolic Androgenic Steroids and Steroid Esters—Comparing Dried Blood Spots Collection Devices and Urine Samples

ABSTRACT

Dried blood spots (DBS) have emerged as a promising complement, and in some settings, an alternative, to urine for anabolic androgenic steroid (AAS) testing, offering advantages such as minimal invasiveness, simplified storage, and transportation. This study evaluated two DBS collection devices—cellulose‐based Capitainer‐B50 and polymer‐based Tasso‐M20—and compared results with traditional urine analysis. Ten self‐reported AAS users were recruited and provided matched urine and DBS samples. High agreement between the two DBS devices was observed, although Capitainer‐B50 showed a slightly greater detection rate, likely due to a higher sample volume (50 μL vs. 17.5 μL) improved analyte recovery, and lower background noise. Notably, DBS enabled detection of testosterone use in all 10 participants, while urine testing missed two cases with naturally low urinary testosterone/epitestosterone (T/E) ratios (most likely UGT2B17 del/del carriers). Moreover, the differentiation between prescribed and illicit use of testosterone esters was also possible in DBS, but not in urine testing, while nandrolone detection in DBS was limited at low concentrations. The findings support DBS as a sensitive and practical tool for AAS detection and provide critical advantages in detecting doping with testosterone esters in individuals with prescribed testosterone therapy and in UGT2B17 deletion carriers.

DBS (Capitainer‐B50 and Tasso‐M20) enabled detection of testosterone esters in all participants, whereas urine testing (T/E) missed cases including potential UGT2B17 del/del carriers and individuals with testosterone prescription. More AAS and steroid‐esters could be detected in Capitainer‐B50 versus Tasso‐M20.

1. Introduction

Dried blood spots (DBS) have been used as an analytical sample matrix for doping tests since 2021. DBS offers some advantages compared to serum samples, including the absence of a need for a phlebotomist, low invasiveness, simplified transportation, easy storage, and high stability of analytes [1, 2]. There are numerous approved sampling devices on the market today, with different absorbent materials such as cellulose paper or polymer based. The extraction and recovery of analytes can vary depending on the absorbent material used in the DBS device. Most sampling devices are designed to collect blood from the fingertip, for example, Mitra, Capitainer, and Hema Spot HF, whereas the Tasso‐M20 device collection is done from the upper arm. In some cases, the pharmacokinetics of drugs may differ between venous and capillary blood [3], and there may also be differences in endogenous concentrations between capillary blood obtained from the fingertip and the upper arm [4]. Therefore, it is important to validate different sampling devices using authentic samples.

Several studies have demonstrated the suitability of DBS as a matrix for detecting testosterone (T) esters. After administration of Sustanon (T‐decanoate, T‐isocaproate, T‐phenylpropionate, and T‐propionate), these esters could be detected in DBS for up to 14 days, depending on the length of the esters [5]. It has been demonstrated that therapeutic doses of T‐undecanoate can be detected in DBS with sensitivity comparable to urinary T/epitestosterone (T/E) [6]. In certain cases, identifying T esters in DBS may eliminate the need for isotope ratio mass spectrometry (IRMS) to confirm T doping. It has also been demonstrated that exogenous steroids such as nandrolone can be detected in DBS with high agreement with urine test results [7]. These findings have resulted in a complete switch from urine to DBS testing in fitness centers in Denmark, considering the many advantages DBS offers in these settings.

Since AAS is prohibited by law in some countries including the Nordic countries, testing for AAS will also be done for legal purposes, such as forensic investigation and workplace testing, as well as diagnostic purposes in healthcare settings. The AAS test panel covered in methods employed outside WADA accredited laboratories is less comprehensive compared to the AAS covered by the WADA prohibited list [8]. Moreover, higher T/E thresholds are applied since IRMS is not usually used for confirmation. Consequently, many samples may be reported as false negatives [9]. Direct targeting of T esters in DBS has high potential to become a complementary method and thereby increase the detection rate of illicit T use.

The objective of this study was to evaluate the performance of two DBS collection devices, Tasso‐M20 and Capitainer‐B50, in detecting various steroid esters in male AAS users. Moreover, the presence of T‐esters and other AAS in DBS samples was compared with the detection in urine to assess the applicability of different micro‐sampling devices.

2. Materials and Methods

2.1. Study Population

Ten male patients declaring the use of AAS when seeking medical care at St Göran Hospital, Department of Medicine, were recruited for this study. Inclusion criteria were an ongoing AAS use (within the last 2 months) and an age between 18 and 65 years. Ethical approval was granted from Swedish Etikprövningsmyndigheten (DNR‐2024‐02827‐01), and the participants received both oral and written information to which they gave their consent. They were asked what doping substances they were using (or have been using) within the last 2 months. DBS samples with both Tasso‐M20 (Tasso Inc., USA) and Capitainer‐B50 (Capitainer, Sweden) and urine were collected at the visit. The DBS samples were shipped to the Helsinki Doping Control Laboratory at room temperature for steroid ester analyses, and urine samples were stored at −20°C prior to analyzes at Stockholm Doping Control Laboratory.

2.2. DBS Analyses

2.2.1. Chemicals and Reagents

d₃‐T and d4‐T undecanoate used as internal standards (IS) were from Cerilliant (Texas, USA) and Alsachim (Illkirch, France), respectively. T esters (acetate, benzoate, caproate, cypionate, decanoate, enanthate, isocaproate, phenylpropionate, propionate, and undecanoate), as well as trenbolone esters (acetate, enanthate), nandrolone esters (decanoate, laurate, undecanoate), ethisterone, methandienone, methenolone, and methyltestosterone, were from Steraloids (Newport, USA). Testosterone hexahydrobenzoate and valerate and nandrolone phenylpropionate, as well as oxymesterone, calusterone, methasterone, gestrinone, stanozolol, trenbolone, fluoxymesterone, and nandrolone, were from TRC (Ontario, Canada). Drostanolone, tetrahydrogestrinone, and methyldienolone were from NMI (Pymble, Australia). Chlorodehydromethyltestosterone was from Cayman Chemical (Michigan, USA) and drostanolone propionate and methenolone acetate from Dr. Erhenstorfer (Augsburg, Germany). The Girard P reagent was purchased from Biosynth (Compton, UK). Acetic acid was from Merck (Darmstadt, Germany). The ultra‐purified water used was of Milli‐Q‐grade (Millipore, Bedford, USA). Methanol (MeOH) was from J.T.Baker (New Jersey, USA), formic acid from Honeywell (Seelze, Germany), and ammonium formate from Fluka (Gillingham, UK).

2.2.2. Sample Preparation

The DBS samples were analyzed using the accredited method used for routine DBS samples. The method has been described previously [10]. Briefly, the whole DBS spots were punched out (cellulose cards) or taken out using tweezers (polymer‐based DBS), transferred into borosilicate tubes, and fortified with 500 μL of methanol containing the ISs (d₃‐T 1 ng/mL and d₄‐T undecanoate 0.5 ng/mL). The spots were extracted using an ultrasonic bath for 15 min. The supernatant was transferred to a new borosilicate tube and evaporated to dryness at 60°C. The dry residue was then derivatized by adding 50 μL of 10% acetic acid in (MeOH) and 20 μL of Girard P reagent (4 mg/mL in H₂O), after which the sample was incubated at 60°C for 10 min. After derivatization, the samples were directly injected into the LC–MS/MS system. The DBS method has been validated for cellulose‐based (Tasso) and polymer‐based (Hemaxis, DBS System SA, Switzerland) devices. The limits of detection (LODs) are listed in the Supporting Information S1.

2.2.3. LC–MS/MS Analysis

An 1290 Infinity autosampler and 6460 Triple Quadrupole, from Agilent, was used. The column was an Agilent Zorbax Eclipse Plus C18 (2.1 × 50 mm, 1.8 μm) operated at 55°C. The mobile phase (A: 2.5‐mM ammonium formate, 0.1% formic acid [FA], B: 2.5‐mM ammonium formate, 0.1% FA in 95% MeOH) flow rate was 0.3 mL/min and injection volume 5 μL. All compounds were monitored in positive ionization (SRM mode). The capillary voltage, drying gas flow, drying gas temperature, nebulizer, nozzle voltage, sheath gas flow, and sheath gas temperature were set at 3500 V, 8 L/min, 200°C, 35 psi, 500 V, 12 L/min, 350°C. To wash out remaining Girard P reagent, the eluent was directed to waste for the first 3 min. The mass spectrometric parameters are presented in Table 1.

TABLE 1.

LC–MS/MS parameters for the analyzed compounds.

Compound	Tgt	ce (V)	Q1	ce (V)
Calusterone	450.3 → 151.0	46	450.3 → 108.1	60
Dehydrochloromethyltestosterone	335.2 → 317.2	8	335.2 → 155.0	36
Drostanolone	438.3 → 359.1	30	438.3 → 331.0	46
Drostanolone propionate	494.3 → 415.1	38	494.3 → 387.0	34
Ethisterone	446.3 → 367.1	34	446.3 → 339.1	30
Fluoxymesterone	470.3 → 391.1	34	470.3 → 363.0	34
Gestrinone	442.2 → 278.0	40	442.2 → 320.0	40
Metandienone	301.2 → 121.0	28	301.2 → 149.0	12
Methasterone	452.3 → 345.0	34	452.3 → 373.1	40
Methenolone	436.3 → 357.1	30	436.3 → 329.1	34
Methenolone acetate	478.3 → 399.2	30	478.3 → 187.0	34
Methyldienolone	420.3 → 313.1	30	420.3 → 341.1	30
Methyltestosterone	436.3 → 357.1	34	436.3 → 329.1	38
Nandrolone	408.3 → 329.2	30	408.3 → 301.2	34
Nandrolone decanoate	562.4 → 483.3	38	562.4 → 455.0	40
Nandrolone laurate	590.4 → 511.3	38	590.4 → 163.1	58
Nandrolone phenylpropionate	540.3 → 461.3	54	540.3 → 105.0	38
Nandrolone undecanoate	576.4 → 497.3	58	576.4 → 469.3	42
Oxandrolone	307.2 → 271.1	12	307.2 → 229.0	15
Oxymesterone	452.3 → 373.0	34	452.3 → 345.0	34
Stanozolol	329.3 → 95.1	48	329.3 → 121.0	44
Testosterone acetate	464.3 → 385.2	48	464.3 → 151.0	44
Testosterone benzoate	526.3 → 447.1	34	526.3 → 419.1	34
Testosterone caproate	520.4 → 441.2	34	520.4 → 151.0	50
Testosterone cypionate	546.4 → 163.0	52	546.4 → 151.0	42
Testosterone decanoate	576.4 → 497.3	58	576.4 → 469.3	42
Testosterone enanthate	534.4 → 455.3	34	534.4 → 427.0	36
Testosterone hexahydrobenzoate	532.4 → 453.2	38	532.4 → 425.2	38
Testosterone isocaproate	520.4 → 441.2	34	520.4 → 151.0	50
Testosterone phenylpropionate	554.3 → 475.4	34	554.3 → 105.0	60
Testosterone propionate	478.3 → 399.2	46	478.3 → 151.0	30
Testosterone undecanoate	590.4 → 511.3	38	590.4 → 163.1	58
Testosterone valerate	506.3 → 427.2	34	506.3 → 399.2	38
Tetrahydroestrinone	446.3 → 339.1	26	446.3 → 367.1	26
Trenbolone	404.2 → 297.1	30	404.2 → 325.1	26
Trenbolone acetate	446.2 → 367.2	26	446.2 → 264.1	38
Trenbolone enanthate	516.3 → 437.1	30	516.3 → 264.1	42
ISTD testosterone‐d3	425.3 → 346.2	46	425.3 → 151.1	34
ISTD testosterone undecanoate‐d4	594.5 → 515.4	42	594.5 → 320.0	54

Abbreviations: ce(V), collision energy (volts); Q1, qualifier transition; Tgt, target transition.

The DBS samples were extracted according to the protocol for cellulose‐based and polymer‐based samples. The DBS samples were classified as positive when an exogenous AAS and steroid ester were detected, and the retention time (RT) and ion ratios were within the limits of TD2023IDCR, regardless of concentrations. Negative and positive QC samples (spiked at a concentration of 2 ng/mL) were included in the analytical batch. The samples were not subjected to further confirmation analyses.

2.3. Urine Analyses

GC–MS/MS (and UHPLC‐HRMS for stanozolol and trenbolone metabolites, respectively) methods were used to analyze for exogenous AAS and T/E. The methods have been described previously [8, 11]. Briefly, 2‐mL urine was hydrolyzed with β‐glucuronidase from E. coli in a 1.1‐M phosphate buffer (pH 7.0) at 50°C for 60 min prior to sample clean‐up. For GC–MS/MS analysis, samples were made basic with the addition of 250‐μL 20% K₂CO₃(aq) and a small scoop of 0.25 g Na₂SO₄(s) prior to liquid–liquid extraction (LLE) with 5‐mL methyl tert‐butyl ether (TBME). Samples were then shaken on a shaker table for 10 min and centrifuged (800 G) for 5 min before placed in an ethanol freeze bath (−24°C). The ether phase was transferred into a new test tube and evaporated under a gentle stream of heated nitrogen gas (80°C). Finally, samples were derivatized with a N‐trimethylsilyl‐N‐methyl trifluoroacetamide mixture and derivatized for 30 min in an 80°C heating block.

For UHPLC‐HRMS analysis, samples were made acidic with the addition of 100‐μL 100% FA and applied on strong cation exchange cartridges for a solid phase extraction (HR‐XC SPE, Chromabond 60 mg/3 mL). Prior to sample application, the cartridges had been activated and conditioned with 0.5‐mL MeOH and 0.5‐mL 2% FA, respectively. Three washing steps followed (1‐mL 2% FA, 1‐mL Milli‐Q water, and 0.5‐mL 10% MeOH) before samples were eluted with 2‐mL 5% NH₃(aq) in 50:50 acetonitrile: MeOH. The eluate was evaporated under a gentle stream of nitrogen gas in a 60°C heating block before being reconstructed in 0.2‐mL 10% MeOH and 50 μL of the original urine.

In results evaluation, thresholds from laboratories employed outside WADA accreditations were applied. Samples with an exogenous AAS above 10 ng/mL were considered as positive, though nandrolone metabolite (19‐NA) and boldenone have higher thresholds of 15 and 30 ng/mL, respectively [12, 13]. For details regarding which AAS and metabolites were included in the analysis, see Table 2. For endogenous AAS, a T/E above 10 was considered as positive for T doping.

TABLE 2.

AAS and the metabolites detected in urine using GC‐MSMS and LC‐MSMS.

Name	IUPAC
Boldenone	17b‐Hydroxy‐androsta‐1,4‐dien‐3‐one
Boldenone M	17b‐Hydroxy‐5b‐androst‐1‐en‐3‐one
DHCMT M	6b‐Hydroxy‐4‐chlorodehydromethyltestosterone
Drostanolone	17b‐Hydroxy‐2a‐methyl‐5a‐androstan‐3‐one
Drostanolone M	3a‐Hydroxy‐2a‐methyl‐5a‐androstan‐17‐one
Epitestosterone
Mesterolone M	3a‐Hydroxy‐1a‐methyl‐5a‐androstan‐17‐one
Mesterolone M2	1a‐Methyl‐5a‐androstane‐3a,17b‐diol
Mesterolone M3 (LTM)	1a‐Methyl‐5a‐androstan‐3,6,16‐triol‐17‐one
Metandienone M	17b‐Methyl‐5b‐androst‐1‐ene‐3a,17a‐diol
Metandienone M2	6b,17b‐Dihydroxy‐17a‐methylandrost‐1,4‐dien‐3‐one
Metandienone M3 (LTM)	17b‐Hydroxymethyl,17a‐methyl‐18‐nor‐androst‐1,4,13‐trien‐3‐one
Metenolone	17b‐Hydroxy‐1‐methyl‐5a‐androst‐1‐en‐3‐one
Metenolone M	3a‐Hydroxy‐1‐methylene‐5a‐androstan‐17‐one
Metenolone M2 (LTM)	16a‐Hydroxy‐1‐methyl‐5a‐androst‐1‐en‐3,17‐dione
Methyltestosterone M	17a‐Methyl‐5a‐androstan‐3a,17b‐diol
Methyltestosterone M2	17a‐Methyl‐5b‐androstan‐3a,17b‐diol
Nandrolone M	19‐Norandrosterone
Nandrolone M2	19‐Noretiocholanolone
Oxandrolone	17b‐Hydroxy‐17a‐methyl‐2‐oxa‐5a‐androstan‐3‐one
Oxandrolone M	17a‐Hydroxy‐17b‐methyl‐2‐oxa‐5a‐androstan‐3‐one
Oxandrolone M2	17b‐Hydroxymethyl‐17a‐methyl‐18‐nor‐2‐oxa‐5a‐androst‐13‐en‐3‐one
Oxandrolone M3	17a‐Hydroxymethyl‐17b‐methyl‐18‐nor‐2‐oxa‐5a‐androst‐13‐en‐3‐one
Oxymesterone	4,17b‐Dihydroxy‐17a‐methylandrost‐4‐en‐3‐one
Oxymetholone M2	2a‐Hydroxymethyl‐17a‐methyl‐5a‐androstane‐3a,6b,17b‐triol
Oxymetholone M3	18‐Nor‐2z,17b‐hydroxymethyl‐17a‐methyl‐5a‐androst‐13‐en‐3a‐ol
Oxymetholone M4 (LTM)	18‐Nor‐17b‐hydroxymethyl‐17a‐methyl‐2a‐methyl‐5a‐androst‐13‐en‐3‐one
Stanozolol M	3′‐Hydroxy‐stanozolol
Stanozolol M2	16b‐Hydroxy‐Stanozolol
Stanozolol M3	17‐Epistanozolol‐N‐glucuronide
Testosterone
Trenbolone	17β‐Hydroxy‐estr‐4,9,11‐trien‐3‐one
Trenbolone M	17α‐Hydroxy‐estr‐4,9,11‐trien‐3‐one

3. Results and Discussion

3.1. Study Population

The participants were males, aged 27–60 years and active gym goers (training > 4 times/week). One participant (S2) was excluded from the study due to the absence of self‐reporting AAS use; instead, he reported the use of growth hormone. All remaining participants declared AAS use within the preceding 2 months, with durations of active AAS use ranging from 1 to 9 years. The doping substances declared by the participants are shown in Table 3. The most common self‐reported AAS was T, declared by 9 of the 10 participants. This agrees with other studies where T has been reported as the most popular AAS among gym goers. Four participants reported the use of only T, without any additional exogenous AAS. Three of these individuals (S6, S9, and S11) stated that they were undergoing medically prescribed T replacement therapy. Two participants received intramuscular injection of T‐undecanoate (Nebido) while one was treated with transdermal T‐gel (Testrox). T‐undecanoate and T‐gel are the two T preparations available on the Swedish market, unless prescribed on license. Notably, the three participants reporting T substitute treatment also stated an illicit use of other T preparations. The use of non‐prescribed (illicit) T, alongside medically prescribed T therapy, is an emerging concern. Although frequently discussed in online forums, the prevalence of this practice remains unknown. The primary motivation behind such “side‐use” is to achieve higher T doses than those prescribed, while retaining a legitimate prescription as a potential justification in the event of a positive doping test.

TABLE 3.

Summary of which AAS each participant reported to use, followed by the detection of AAS in different matrix tested.

Test subject	Declared by test subject	Substances found in any analysis	Tasso‐M20 (estimated concentrations—ng/mL)	Capitainer‐B50 (estimated concentrations—ng/mL)	Urinary AAS (ng/mL)
S1	3 T‐esters	Drostanolone + M	—	—	44 + 1253
		Testosterone	—	—	7.1 (T/E 2.31)
		Testosterone decanoate		0.1	—
S3		Methandienone	32	37	—
		Methandienone M + M2 + M3 (LTM) + M4	—	—	181 + 140 + 1105 + 178
	Testosterone	Stanozolol M + M2	—	—	0.16 + 0.21
	Trenbolone	Testosterone		—	412 (T/E 137)
	Methenolone enanthate	Testosterone enanthate	0.6	10	—
		Trenbolone enanthate		0.7	—
		Trenbolone	2.8	3	20.2
S4		Drostanolone + M	—	—	0.5 + 3.9
		Testosterone	—	—	243 (T/E 74)
	Testosterone	Testosterone isocaproate	0.5	1	—
		Testosterone phenylpropionate	0.2	0.3	—
		Testosterone decanoate	1.1	0.2	—
		Testosterone propionate		0.1	—
S5		Metenolone + M + M2	—	—	5.3 + 4.3 + detected
		Nandrolone	—	0.2	—
		Nandrolone phenylpropionate	1.2	6	—
		Testosterone	—	—	1.3 (T/E 0.47)
	Testosterone	Testosterone isocaproate		1.6	—
	Methenolone enanthate	Testosterone cypionate	—	3.8	—
		Testosterone decanoate	—	0.4	—
		Testosterone enanthate	1.5	9	—
		Testosterone phenylpropionate	0.1	0.7	—
		Testosterone propionate	0.1	0.3	—
S6	Testosterone undecanoate a	Testosterone	—	—	45 (T/E 37)
	Short T‐esters	Testosterone undecanoate	0.2	0.1	—
		Testosterone propionate	0.2	0.1	—
S7		Boldenone + M	—	—	17 + 3.9
		Drostanolone + M	—	—	3.4 + 91
		Methandienone	—	0.4	—
		Metandienone M + M2 + M3 (LTM)	—	—	7.5 + 0.8 + 1626
	Trenbolone	Nandrolone M + M2	—	—	4.2 + 1.7
	Metenolone enanthate	Testosterone	—		6.1 (T/E 4.17)
		Testosterone isocaproate	—	0.06	—
		Testosterone phenylpropionate	—	0.06	—
		Testosterone propionate	—	0.3	—
		Trenbolone enanthate	0.1	1.7	—
		Trenbolone	5.4	8.6
		Trenbolone acetate	—	1	71
S8		Boldenone + M	—	—	244 + 66
	Testosterone a	Drostanolone + M	—	—	9.5 + 172
	Nandrolone decanoate	Testosterone	—	—	217 (T/E 66)
	Drostanolone propionate	Testosterone enanthate	0.1	1.2	—
		Nandrolone	0.5	0.2	—
		Nandrolone M + M2	—	—	930 + 115
S9	Testosterone	Testosterone	—	—	6.1 (T/E 9.7)
	Other T‐esters	Testosterone propionate	—	0.9	—
S10		Boldenone + M	—	—	6.7 + 3.4
		Nandrolone	2	1.3	—
	Testosterone	Nandrolone M + M2	—	—	5686 + 1138
	Nandrolone decanoate	Nandrolone phenylpropionate	0.9	1.9	—
		Testosterone	—	—	702 (T/E 73)
		Testosterone cypionate	—	1.3	—
		Testosterone enanthate	1	2.2	—
S11		Drostanolone + M	Tasso N/A	—	9.4 + 99
		Testosterone		—	296 (T/E 93)
	Nandrolone decanoate	Testosterone decanoate		0.2	—
	Testosterone propionate	Testosterone isocaproate		0.8	—
	Testosterone phenylpropionate	Testosterone phenylpropionate		0.3	—
	Testosterone isocaproate	Testosterone propionate		0.5	—
	Testosterone decanoate	Testosterone enanthate		3.5	—
	Testosterone undecanoate a	Nandrolone		0.3	—
		Nandrolone M + M2		—	1294 + 158
		Nandrolone phenylpropionate		1.9	—
		Oxandrolone + M		41	1097 + 633

Abbreviations: LTM = long‐term metabolite; M = metabolite, see Table 2 for details; N/A = not available.

^a Prescribed for testosterone replacement therapy.

3.2. Urine Analyses

The results from urine analyses are presented in Table 3. Most participants tested positive for T, followed by drostanolone. In 80% of the urine samples, one or more AAS not declared by the users were detected. This discrepancy aligns with findings from previous studies, in which the presence of AAS in urine did not always correspond with self‐reported use [14, 15]. Possible explanations include contaminations, the presence of fortified AAS, or inaccurately labeled products [16, 17]. Many gymgoers consume dietary supplements, which are known to be a risk of AAS contamination [18]. The two most frequently detected non‐declared AAS in urine were drostanolone and boldenone. Notably, drostanolone was also the most commonly undeclared AAS in a previous study of Swedish AAS users [15].

3.3. DBS Analyses—Comparison Tasso and Capitainer

The LOD from the validation has been presented in [10] and summarized in Table S1. Minor differences in the LODs were observed between polymer and cellulose‐based DBS. Overall, the LODs fall within the range reported in previous DBS studies and are considered sufficient for the detection of steroid esters [7, 19]. One limitation is that the validation of the cellulose‐based DBS was performed with HemaXis (10 μL) and not Capitainer‐M50 (50 μL). However, it is reasonable to assume that larger blood volumes might result in even lower LODs. The aim of this study was to evaluate the performance of different DBS collection devices in samples obtained from individuals actively using AAS. Such samples may be more complex than those from athletes, as AAS users often use several additional drugs, both therapeutically and illicitly, as well as various supplements. Moreover, AAS use is associated with elevated lipid and hematocrit values. These factors together may contribute to ion suppression in the LC‐MSMS analysis, reduce analyte recovery, and thereby influence the LOD. Therefore, comparing the detection of steroid esters in polymer versus cellulose‐based DBS collected from AAS‐using individuals is of particular interest.

The detection of AAS using the Tasso‐M20 and Capitainer‐B50 samples is shown in Table 3. The levels of the different esters detected in both devices showed significant correlations (Spearman rank test r _s  = 0.57, p < 0.01). A Bland–Altman analysis showed a calculated bias of −1.7 ng/mL (Figure 1). Due to the limited number of data points for each steroid ester, it is not possible to compare the DBS sampling devices for specific analytes. However, since the method is not intended for absolute quantification of the esters, as their detection indicates illegal use regardless of concentration, this minor difference is not of any clinical concern. In some cases, however, a greater number of steroid esters were identified in Capitainer‐B50 samples compared to Tasso‐M20. Many of these esters were detected at low levels in Capitainer‐B50, and their absence in the Tasso‐M20 samples is probably attributable to the smaller sample volume: 17.5 μL for Tasso‐M20 versus 50 μL for Capitainer‐B50. Still, Tasso‐M20 failed to detect levels > 1 ng/mL of short‐acting T esters (T‐cypionate) in two subjects (S5 and S10.) The reason some T esters were not detected in Tasso‐M20 at high concentrations may be due to low recovery, in line with another study showing low recoveries for the shorter esters with Tasso‐M20 compared to cellulose‐based DBS [20].

FIGURE 1.

Bland–Altman plot comparing Tasso and Capitainer concentrations for various AAS. For improved visualization, the sample with methandienone (> 30 ng/mL) was excluded in the figure.

Nandrolone was more frequently detected in Capitainer‐B50 samples, and in one subject (S7), methandienone was detected in Capitainer‐B50 but not in the corresponding Tasso‐M20. This discrepancy may, in part, be explained by the higher background noise in the polymer‐based Tasso‐M20 compared to the cellulose‐based Capitainer‐B50. A previous study reported that the Tasso‐M20 exhibited RT shifting and chromatographic distortion for most steroid esters in contrast to Mitra [21]. Moreover, analyte extraction from polymer‐based DBS can be challenging, as dried erythrocytes may occlude the pores of the polymer material, thereby reducing extraction efficiency [22].

3.4. Detection in DBS Versus Urine Analyses

Steroid esters were detected in all collected DBS samples. Table 3 provides a comparison of the AAS detected in DBS and urine samples for each participant; a high level of concordance was observed between the two matrices, with most AAS identified in DBS also being detected in urine samples, except S5 and S9 using the thresholds stated above. For the three participants who illicitly used additional T preparations beyond their prescribed therapy, the T/E ratio, although elevated (S6 and S11), would not have indicated doping. One of the potentially dangerous adverse side effects of AAS use is an elevated erythrocyte volume fraction. T is equally potent as other AAS in increasing blood values. It would be both legally and medically justified to implement testing methods capable of differentiating between different T preparations. In all three participants, one or more non‐prescribed T esters were detected in the DBS samples. Subsequently, these individuals would have tested positive only if subjected to a steroid ester analysis. Under the World Anti‐Doping Code, athletes may be granted a therapeutic use exemption for T solely in cases of hypogonadism with a confirmed organic etiology, which is uncommon in otherwise healthy athletes [23]. In such cases, DBS could serve as an important complementary matrix to traditional urine analysis. Participant S5 displayed a low T/E despite the presence of multiple T esters in his DBS sample. He is most likely homozygous for the UGT2B17 gene deletion (del/del), which is associated with low T/E, both at baseline and after T administration [24, 25]. These results highlight the utility of DBS to identify T doping in del/del subjects. The finding is consistent with previous results demonstrating that analyzing T esters in serum increases the chances of detecting prohibited T administration in individuals with low T/E [26]. Another interesting approach could be the analysis of T esters in urine, as low concentrations of T‐propionate and T‐enanthate have been detected in IRMS confirmed T positive urine samples [27].

The sensitivity to detect exogenous AAS in DBS varied among substances. In the case of nandrolone, non‐esterified nandrolone was detected in DBS samples from participants where high urinary 19‐NA (nandrolone M) concentrations were present (930–5688 ng/mL). However, in urine samples with low 19‐NA levels (4.2 ng/mL), nandrolone was not detected in DBS. Notably, in participant S5, nandrolone was detected in DBS (0.2 ng/mL), without the presence of 19‐NA in urine. The reason for this discrepancy between the matrices is not known, but it is possible that he provided samples shortly after nandrolone administration. The nandrolone levels in DBS samples were in a similar range as found in DBS collected from Danish fitness centers [7]. Nandrolone has a relative short half‐life in blood [28], whereas its primary urinary metabolite 19‐NA can be detected over 9 months following administration of nandrolone decanoate [29]. In three DBS nandrolone positive samples, nandrolone phenylpropionate was identified. This ester, commercially known as Durabolin, has a shorter half‐life than nandrolone decanoate (Deca‐durabol). According to disclosures from the participants, the short acting form of nandrolone appears to be more commonly used. Additionally, with DBS it was possible to detect many other exogenous AAS, for example, trenbolone, boldenone, and methandienone.

4. Conclusion

Overall, our results corroborate previous findings that DBS can serve as a complementary or even alternative matrix for AAS detection. The Capitainer‐B50 device demonstrated a higher detection rate of AAS compared to Tasso‐M20. Nonetheless, both DBS sampling devices proved effective for steroid ester analyses, particularly for T detection, and thereby increasing the chances to detect illicit T use.

5. Limitations

One limitation of this study is the small sample size (n = 10), so the comparative detection of AAS should be interpreted with caution. All participants were using T preparations; thus, the main findings were centered around T detection. Due to the legal status of doping in Sweden, recruiting volunteers for scientific studies in this field remains a considerable challenge. Nonetheless, this is the first study to compare the detection of steroid esters using cellulose and polymer‐based DBS devices.

Another limitation of this study is that only men were included. While AAS use is estimated to be approximately 10 times more prevalent among men in the general population, the pattern differs in sports where AAS use is believed to be more evenly distributed between sexes. Many female AAS users, on the other hand, prefer transdermal or oral T preparations over intramuscular injections of steroid esters. Consequently, DBS‐based detection may provide limited additional value in this population. Instead, serum T and Androstenedione (A4) have been shown to be valid markers for T doping in female athletes undergoing longitudinal testing [30]. These markers have also been shown to be quantifiable in DBS samples [31]; however, the applicability of a longitudinal monitoring approach in the general population is limited.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Limit of detection (LOD) for the anabolic androgenic steroids and steroid esters included in the DBS analyses.

Suominen T., von Walden J., Harju L., et al., “Detection of Anabolic Androgenic Steroids and Steroid Esters—Comparing Dried Blood Spots Collection Devices and Urine Samples,” Drug Testing and Analysis 17, no. 12 (2025): 2374–2383, 10.1002/dta.3950.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.