Sixteen apparently healthy, recreationally-active males with a mean age of 26.6 ± 4.9 years, height of 180.2 ± 6.3 cm, body fat of 14.9 ± 4.8 %, and body weight of 87.3 ± 13.2 kg served as participants in the study. All participants were cleared for participation by passing a mandatory medical screening. Participants with contraindications to exercise as outlined by the American College of Sports Medicine and/or who had consumed any nutritional supplements (excluding multi-vitamins) such creatine monohydrate or various androstenedione derivatives or pharmacologic agents such as anabolic steroids two months prior to the study were not allowed to participate. All eligible subjects signed a university-approved informed consent document. Additionally, all experimental procedures involved in this study conformed to the ethical considerations of the Helsinki Code.
Testing sessions were performed at week 0 and after weeks 1, 3, 8, and 11 in which blood and urine samples were obtained and where body composition, serum hormones, blood and urinary clinical safety markers, and systemic hemodynamic safety markers were evaluated.
Body composition assessment
Total body mass (kg) was determined on a standard dual beam balance scale (Detecto Bridgeview, IL). Percent body fat, fat mass, and fat-free mass were determined using DEXA (Hologic 4200 W, Waltham, MA). Quality control calibration procedures were performed on a spine phantom (Hologic X-CALIBER Model DPA/QDR-1 anthropometric spine phantom) and a density step calibration phantom prior to each testing session. Total body water and compartment-specific fluid volumes were determined by bioelectric impedance analysis (Xitron Technologies Inc., San Diego, CA).
Blood and urine collection
Venous blood samples were obtained from an antecubital vein into a 10 ml collection tube. Blood samples were allowed to stand at room temperature for 10 min and then centrifuged. The serum was removed and frozen at -80°C for later analysis. Urine samples were obtained in mid-stream into a collection container using a standard collection protocol. Urine samples were frozen at -80°C for later analysis. Blood and urine samples were obtained after a 12-hour fast and standardized to the same time of day for each sample.
Based on the premise that there were no muscle performance measurements to be made in the study, such as muscle strength and power, that could otherwise generate a so-called "placebo effect," the decision was made not to utilize a placebo group. Participants were equally divided, matched by age and body mass, and then randomly assigned in double-blind fashion to an eight-week supplementation protocol consisting of the daily oral ingestion of either 300 mg or 600 mg of 6-OXO [androst-4-ene-3,6,17-trione (ErgoPharm, Champaign, IL)]. For the 300 mg group, (n = 8; total body mass = 79.3 ± 13.2 kg, fat-free mass = 67.1 ± 7.9 kg; body fat = 14.7 ± 5.4 %) 100 mg was ingested in the morning with breakfast and 200 mg was ingested with the evening meal. For the 600 mg group, (n = 8; total body mass = 81.1 ± 13.3 kg, fat-free mass = 69.0 ± 12.1 kg; body fat = 15.0 ± 4.2 %) 300 mg was ingested in the morning with breakfast and 300 mg was ingested with the evening meal. For days where no exercise occurred, the supplements were ingested in the same timely fashion. After the supplementation period, a three-week washout period was required during which neither supplement was ingested. Upon analysis of serum testosterone from the baseline blood samples at week 0, it was confirmed that all participants completing the study were eugonadal [10–30 nmol/L (27–107 ng/ml)].
Physical activity, dietary intake records, and supplementation compliance
During both the supplementation and washout periods the participants' physical activity and dietary intake were not supervised; however, it was required that all participants keep detailed physical activity and dietary records and not change their routine dietary habits or level of physical activity throughout the course of the study. As such, participants were required to keep weekly physical activity records and four-day dietary records during weeks 0, 1, 3, 8, and 11 and turn in their physical activity and dietary records during each testing session. Each participant returned all of their dietary and physical activity evaluations at the required time points for a 100% compliance rate. The four-day dietary recalls were evaluated with the Food Processor dietary assessment software program (ESHA Research, Salem, OR) to determine the average daily macronutrient consumption of fat, carbohydrate, and protein. In an effort to ensure compliance to the supplementation protocol, participants were supplied with the appropriate number of capsules to be ingested during the time between testing sessions 1, 3, and 8. Upon reporting to the lab at each of the respective testing sessions, participants returned the empty containers and a capsule count was performed if necessary.
Reported side effects from supplements
At the last four testing sessions, participants reported by questionnaire whether they tolerated the supplement, supplementation protocol, as well as report any medical problems/symptoms they may have encountered throughout the study.
Hemodynamic clinical safety markers
At each testing session, participants assumed a supine position for 15 minutes and had their heart rate (HR), systolic blood pressure (SBP), and diastolic blood pressure (DBP) determined to assess the hemodynamic safety of supplementation with 6-OXO. Heart rate was determined by use of a Polar heart rate monitor (Polar, San Ramon, CA), and blood pressure was assessed with a mercurial sphygmomanometer using standard procedures.
Blood and urinary clinical markers
The serum clinical chemistry variables glucose, total protein, blood urea nitrogen, creatinine, BUN/creatinine ratio, uric acid, AST, ALT, CK, LDH, GGT, albumin, globulin, sodium, chloride, calcium, carbon dioxide, total bilirubin, alkaline phosphatase, triglycerides, cholesterol, HDL, and LDL were determined with a Dade Dimension RXL clinical chemistry analyzer (Dade-Behring, Inc., Newark, DE). The whole blood hematological variables, hemoglobin, hematocrit, red blood cell counts, MCV, MCH, MCHC, RDW, neutrophils, lymphocytes, monocytes, eosinophils, and basophils, were determined with an Abbott Cell Dyne 3500 hematology analyzer (Abbott Laboratories, Chicago, IL). The urinary variables glucose, ketones, blood, protein, nitrite, bilirubin, leukocyctes, specific gravity, pH, urobilinogen were analyzed with a Bayer Clinitek 200 Plus urine analyzer (Bayer Diagnostics, Tarrytown, NY).
Serum TT, FT, DHT, estradiol, estrone, estriol, SHBG, LH, growth hormone (GH), cortisol (Diagnostics Systems Laboratories, Webster, TX), and FSH (Alpco Diagnostics, Windham, NH) using enzyme-linked immunoabsorbent assays (ELISA) and enzyme-immunoabsorbent assays (EIA) with a Wallac Victor-1420 microplate reader (Perkin-Elmer Life Sciences, Boston, MA), and the assays were performed at a wavelength of either 450 or 405 nm, respectively. The average specificity for all assays was 3.5 pg/ml, and in all cases the intra-assay and inter-assay variances were < 10%. Additionally, the amount of cross-reactivity between androstenedione and FT, TT, and DHT was 0.06%, 0.09%, and 1.9%, respectively.
Due to the likelihood of the body composition, serum hormones, and serum and urinary clinical chemistry marker dependent variables being related to one another, statistical analyses were performed by utilizing three separate repeated-measures two-factor [treatment groups (2) × time point (5)] mixed methods multivariate analysis of variance (MANOVA). The use of MANOVA also reduces the risk of Type I errors, by controlling for alpha level that could result with the use of repeated analyses of variance (ANOVA). Box M tests were performed to test for differences in covariance matrices. Bartlett's Test of Sphericity were performed to test that the variance and covariance matrix of the dependent variables was circular in form which would allow for accurate interpretation of univariate ANOVAs. Levene's Test of Equality of Error was performed to test for equality of variance for each dependent variable. Further analysis of the main effects for Group and Test were performed by separate one-way ANOVAs and by using Sidak pair-wise comparisons. Significant between-group differences were then determined by employing the Tukey HSD Post Hoc Test when the variances were equal and a Games-Howell Post Hoc Test was used when the variances were not equal. The changes from post- to pre-training for each criterion variable were then analyzed with a one-way ANOVA. Non-significant trends (p < .10) were reported as partial Eta squared to illustrate the effect size, where an effect size of 0.51 is relatively strong, 0.24 is moderate and is 0.17 relatively weak . All statistical procedures were performed using SPSS 13.0 software and a probability level of < 0.05 was adopted throughout.