Confounders of haemoglobin mass in athletes : implications for anti-doping

  • Clare E. Humberstone

Student thesis: By Publication


The Athlete Biological Passport (ABP) is used for the indirect detection of blood doping in athletes. The calculations within the ABP (Adaptive model) monitor longitudinal changes in characteristics of an athlete’s blood, specifically haemoglobin concentration ([Hb]) and percent reticulocytes (%Ret),and ‘flag’ unusual variations that may be characteristic of doping. The inclusion of total haemoglobin mass (Hbmass) as a marker in the ABP may improve the Passport’s sensitivity. However, concerns have been expressed about the suitability of Hbmass measurements for this purpose. This thesis investigated two potential barriers to including Hbmass in the ABP: the lack of a quality control system for Hbmass measurement and the potential confusion between longitudinal Hbmass profiles of doped and non-doped athletes. The use of custom-made quality control solutions eliminated the majority of between laboratory differences in Hbmass measures. Analytical error associated with making successive Hbmass measurements in three different laboratories was reduced from 2.4% to 1.7% when the quality controls were used. These findings demonstrated that using quality control solutions would ensure that Hbmass results from different laboratories were equivalent if Hbmass was included as a marker in the ABP. The effects of various confounders on Hbmass in non-doped athletes were quantified. These confounders were investigated specifically for their potential to increase the biological variability of Hbmass. For detection of blood doping, these confounders represent a ‘worstcase- scenario’ for the variability of Hbmass in non-doped athletes. Ultra-endurance triathlon racing (+3.2%),Classical altitude training (+3.8%) and Live High: Train Low altitude training (two estimates,+4.0% and +4.3%) each caused substantial mean increases in Hbmass. Conversely, reduced training (-2.3%) and surgery (-2.7%) in injured / ill athletes caused substantial mean decreases in Hbmass. Acute Intermittent Hypoxic Exposure did not substantially affect Hbmass (-0.3%). The effects of microdoses of recombinant human erythropoietin (rHuEPO) on Hbmass were also examined to represent a ‘worst-case-scenario’ for the detection of doped athletes using Hbmass in the ABP. Over a 12 week period, rHuEPO microdosing caused a mean increase of Hbmass of 11.0%,with individual responses ranging from increases of 1.4% to 19.2%. Finally, an investigation was carried out to determine which of six different Hbmass Adaptive models might be suitable for inclusion in the ABP. The sensitivities and specificities of these models were compared in a sample of 159 non-doped and 18 doped athletes. In models that used Hbmass as a single marker, the sensitivity and specificity of the model was heavily influenced by the estimate of Hbmass biological within-subject (BioWS) variance included in the calculations. These models were each named after the first author of a publication in which the BioWS variance of Hbmass in athletes was estimated. Due to their low specificities in non-doped athletes, neither of two Hbm (Prommer) models would be suitable for inclusion in the ABP. In contrast, based on specificities close to 100%,any of the Hbm (Pottgiesser),Hbm (Morkeberg) or Hbm (Eastwood) models may be suitable for inclusion, although each model only offered ~20% sensitivity to rHuEPO doping. The novel ONhm+ret model, which combined Hbmass and %Ret into a single marker, would not be useful in the ABP due to its low specificity. Overall ,the inclusion of Hbmass may improve the sensitivity of the ABP, particularly to microdose rHuEPO doping. The sensitivities of the Hbm (Pottgiesser), Hbm (Morkeberg) and Hbm (Eastwood) models should be examined in a larger sample of doped athletes. Unfortunately, these models may be susceptible to recording false-positive results in some extreme cases of Hbmass perturbation in ill or injured athletes.
Date of Award2012
Original languageEnglish
Awarding Institution
  • University of Canberra
SupervisorJudith Anson (Supervisor), Christopher Gore (Supervisor) & Philo Saunders (Supervisor)

Cite this