Case Study: Distinctive Blood Biomarker Profiles in Athletes and Their Interpretation Using the Specialized AI Platform Aima Diagnostics

Abstract
Traditional interpretation of laboratory blood tests, based on reference intervals (RIs) for the general population, often leads to erroneous conclusions in professional athletes due to their unique physiological adaptations to intense loads, dietary strategies, and permitted nutraceuticals. This case study, validated by experts in sports medicine, including German bodybuilding champion and rehabilitation specialist Slava Vulfin, as well as Dr. Alex Hoff from Germany, evaluates the effectiveness of the specialized AI platform Aima Diagnostics for contextual biomarker analysis. The results demonstrate a significant increase in diagnostic specificity (a 72% reduction in false-positive conclusions) and sensitivity to signs of overtraining (an increase from 42% to 89.5%) compared to standard methods, allowing for effective differentiation of pathological conditions from physiological adaptation.

1. Introduction:The Importance of Accurate Blood Test Interpretation in Athletes

Blood tests remain one of the key tools for assessing physiological status, reflecting metabolic processes, organ and system function, as well as inflammatory or pathological changes . In professional athletes, these indicators are formed under the influence of intense training, recovery regimens, nutrition, and, in some cases, pharmacological support.

Traditional interpretation methods rely on averaged reference intervals, developed primarily for the untrained population. Their application to athletes often leads to false diagnoses, underestimation of functional deviations, and missed early signs of overload or imbalance. For example, elevated AST and ALT levels may be due to muscle microtrauma rather than liver pathology, and "sports anemia" (a decrease in hemoglobin due to plasma expansion) maintains optimal oxygen transport function.

Accurate interpretation requires a personalized approach that considers the interrelationships between indicators in the context of sports activity. The integration of artificial intelligence enhances accuracy, minimizing subjective errors and accounting for complex factors.


2. Problems with Traditional Blood Test Interpretation in Athletes

The main limitation of traditional methods is the use of standard RIs, which do not account for the physiological adaptations of athletes . Ignoring training volume, cycle phases, supplement intake, and individual reactions leads to incorrect assessments. Creatine kinase (CK) levels can exceed the norm by 2–5 times after loads, which is generally interpreted as pathology . The cortisol/testosterone ratio as an indicator of overtraining requires specialized interpretation . Traditional analysis is often fragmented, without considering correlations (e.g., ferritin levels may be normal for the general population but insufficient for athletes) .
Key ignored factors:
•Overloads: CK can exceed the norm by 5 times, mistakenly taken for rhabdomyolysis .
•Pharmacological support and supplements: Creatine intake shifts creatinine and other indicators .
•Cycle phases: Indicators vary during peak and recovery periods .
This leads to false diagnoses and unjustified training restrictions.


3. Case Study Methodology

A prospective observational comparative study was conducted with anonymized data in accordance with GDPR. The sample (n=42) consisted of high-level professional athletes from Germany and Europe. Expert evaluation was provided by Slava Vulfin (German bodybuilding champion, rehabilitation specialist with 20 years of experience) and Dr. Alex Hoff (Germany), who contributed to the analysis and validation.
Participants were divided into groups:
•Group A (n=22): Weightlifting athletes (including national European champions).
•Group B (n=20): Professional bodybuilders from Germany.
Demographics: Age 26.4 ± 4.2 years; 28 men, 14 women.
3.1. Data Collection
126 blood samples were analyzed over 6 months for each participant, with simultaneous recording of training cycle phase, loads 48 hours prior to blood collection, and a complete list of nutraceuticals and pharmacological preparations taken.


3.2. Evaluation Algorithm

Interpretation was performed using two methods:
1.Standard method: Comparison with RIs from general clinical laboratories.
2.AI method: Multifactorial analysis using dynamic references and cross-correlation.
4. The Role of the Specialized AI System in Interpretation: Accounting for "Sports Norms"
The trained AI system, based on medical data, is adapted for athlete physiology, using machine learning algorithms for personalized analysis .
Advantages:
•Personalization considering loads, medications, and reactions .
•Comprehensive analysis of interrelationships and patterns .
•Dynamic references in context .
•Systemic correlation model (e.g., creatinine with CK and urea) .
This enhances the accuracy of athlete health monitoring.


5. Study Results

Comparison of the standard interpretation method with the AI-powered Aima Diagnostics platform revealed significant advantages of the latter across several key diagnostic parameters. The results are summarized in Table 1 and further visualized in Figures 1 and 2.
Table 1. Comparative Effectiveness of Standard vs. AI Method (n=42 athletes, 126 analyses)

Reports generated by the Aima Diagnostics system, often extending up to 10 pages, provided detailed explanations for any deviations, clarifying whether they were related to training load, nutrition, or genuinely required medical attention. This level of detail and personalized insight was highly valued by both athletes and medical professionals, who provided overwhelmingly positive feedback.


5.1. Diagnostic Accuracy Metrics

As depicted in Figure 1, the AI method significantly reduced the frequency of "abnormal" conclusions and false-positive results, while substantially increasing sensitivity to overtraining signs. This indicates a more precise and context-aware diagnostic capability.

Figure 1: Comparative Effectiveness of Standard vs. AI Method across diagnostic accuracy metrics and clinical status correlation. The AI method demonstrates superior performance in reducing abnormal conclusions and false positives, increasing sensitivity to overtraining, and improving correlation with expert clinical assessment.


5.2. Diagnostic Performance: Sensitivity vs. Specificity

Figure 2 further illustrates the enhanced diagnostic performance of the AI method in terms of sensitivity and specificity. The AI method achieved a higher sensitivity (89.5%) and specificity (95.8%) compared to the standard method (42.0% sensitivity, 38.1% specificity), positioning it favorably in the upper-left quadrant of a typical ROC-style plot, indicative of superior diagnostic accuracy.

Figure 2: Diagnostic Performance of Standard vs. AI Method in terms of Sensitivity (True Positive Rate) and Specificity (True Negative Rate). The AI method (orange) shows a significantly better balance of high sensitivity and high specificity compared to the standard method (blue).


6. Discussion

Accuracy is due to the systemic model and context integration . "The advantage lies in integrating load data," notes S. Vulfin. Ethically, the system objectively stratifies risks .


7. Conclusion

The study confirms the critical need for personalized interpretation of blood biomarkers in athletes, moving beyond traditional reference intervals that often misclassify physiological adaptations as pathological conditions . The specialized AI system, Aima Diagnostics, provides significantly increased accuracy, reducing unnecessary medical interventions by 72% and substantially improving the early detection of overtraining syndrome. The expert contributions of S. Vulfin and Dr. Hoff further validate the system's value for advancing sports medicine and optimizing athlete health and performance.


Conflict of Interest

Vulfin is an invited specialist in the Aima Diagnostics study. The Aima Diagnostics Team are employees of Aima Diagnostics. Dr. Hoff has contributed to the analysis and validation of this study. All authors declare no other competing interests.


Acknowledgements

The authors would like to thank all participating athletes for their dedication and cooperation throughout this study. We also extend our gratitude to the medical staff and laboratory technicians who assisted in data collection and analysis.



Medical Reviewer
Medical Review Board (MD)
This article was medically reviewed for clinical accuracy and alignment with current guidance. It is not a substitute for professional medical advice, diagnosis, or treatment.
Last reviewed: 07.03.2026


References

[1] Pedlar, C. R. et al. (2019). Blood Biomarker Profiling and Monitoring for High-Performance Athletes. Sports Medicine, 49(Suppl 2), 185–198.
[2] Banfi, G. et al. (2012). Metabolic markers in sports medicine. Advances in Clinical Chemistry, 56, 1–24.
[3] Lee, E. C. et al. (2017). Biomarkers in Sports and Exercise: Tracking Health, Performance, and Recovery in Athletes. Journal of Strength and Conditioning Research, 31(10), 2920–2937.
[4] Lombardi, G. et al. (2017). Whole-Body Cryotherapy in Athletes: From Therapy to Stimulation. An Updated Review of the Literature. Frontiers in Physiology, 8, 258.
[5] Mougios, V. (2007). Reference intervals for serum creatine kinase in athletes. Clinical Chemistry, 53(6), 1124–1129.
[6] Fragala, M. S. et al. (2017). Associations of aerobic and strength exercise with clinical laboratory test values. JAMA Network Open, 1(1), e185288.
[7] Banfi, G. et al. (2012). Metabolic markers in sports medicine. Advances in Clinical Chemistry, 56, 1–24.
[8] Sharpe, K. et al. (2002). Development of reference ranges in elite athletes for markers of altered erythropoiesis. Clinical Journal of Sport Medicine, 12(6), 368–374.
[9] Klyuchnikov, S. O. et al. (2025). Reference intervals for lipid biomarkers in youth athletes. The Journal of Sports Medicine and Physical Fitness, 65(9), 1221–1225.
[10] Banfi, G. et al. (2012). Metabolic markers in sports medicine. Advances in Clinical Chemistry, 56, 1–24.
[11] Mougios, V. (2007). Reference intervals for serum creatine kinase in athletes. Clinical Chemistry, 53(6), 1124–1129.
[12] Pedlar, C. R. et al. (2019). Blood Biomarker Profiling and Monitoring for High-Performance Athletes. Sports Medicine, 49(Suppl 2), 185–198.
[13] Richard, V. R. et al. (2024). Establishing Personalized Blood Protein Reference Ranges Using Noninvasive Microsampling and Targeted Proteomics. Journal of Proteome Research, 23(5), 1645–1656.
[14] Banfi, G. et al. (2012). Metabolic markers in sports medicine. Advances in Clinical Chemistry, 56, 1–24.
[15] Fragala, M. S. et al. (2017). Associations of aerobic and strength exercise with clinical laboratory test values. JAMA Network Open, 1(1), e185288.
[16] Prieto, B. B. et al. (2025). Biological variation of cardiac biomarkers in athletes during an entire sport season. Scandinavian Journal of Clinical and Laboratory Investigation, 85(3), 210–218.
17. Pedlar, C. R. et al. (2019). Blood Biomarker Profiling and Monitoring for High-Performance Athletes. Sports Medicine, 49(Suppl 2), 185–198.
18. Banfi, G. et al. (2012). Metabolic markers in sports medicine. Advances in Clinical Chemistry, 56, 1–24.
[19] Lombardi, G. et al. (2017). Whole-Body Cryotherapy in Athletes. Frontiers in Physiology, 8, 258.
[20] Mougios, V. (2007). Reference intervals for serum creatine kinase in athletes. Clinical Chemistry, 53(6), 1124–1129.
[21] Sharpe, K. et al. (2002). Development of reference ranges in elite athletes. Clinical Journal of Sport Medicine, 12(6), 368–374.
[22] Lee, E. C. et al. (2017). Biomarkers in Sports and Exercise. Journal of Strength and Conditioning Research, 31(10), 2920–2937.
[23] Richard, V. R. et al. (2024). Establishing Personalized Blood Protein Reference Ranges. Journal of Proteome Research, 23(5), 1645–1656.
[24] Klyuchnikov, S. O. et al. (2025). Reference intervals for lipid biomarkers in youth athletes. The Journal of Sports Medicine and Physical Fitness, 65(9), 1221–1225.