This study has gathered the largest sample so far of amateur elite boxers and matched controls to examine a possible relationship between MTBI in amateur boxing and CSF biomarkers.
Although only one of the boxers had self-reported symptoms of a possible concussion and the clinical examination was normal in all the boxers, the data demonstrate that concentrations of NFL, GFAP, S-100B and T-tau in CSF were increased within 6 days after a bout in more than 80% of the amateur boxers indicating acute axonal and neuronal damage. NFL, GFAP and T-tau are specific markers for damage of the central nervous system and increased concentrations of NFL 
and GFAP 
have previously been found both after acute and chronic brain injuries caused by different types of trauma. Both NFL and GFAP further remained significantly elevated after a resting period of at least 14 days (test B) in the boxers compared to the controls in the present study. Most of the boxers with increased NFL and GFAP concentrations at test B had fought many bouts, both before test A, and between the two tests, which may have resulted in a cumulative effect. One previous study has analysed CSF biomarkers in relation to boxing, where 14 boxers compared to 10 controls showed elevated concentrations of NFL, GFAP and T-tau 7–10 days after a bout. Only NFL remained elevated at a 3 month follow-up and correlation was found between CSF concentration of NFL and an injury severity score 
. Also in the present study, a correlation was found between a composed boxing exposure index and NFL, but construction of such an index is difficult since the development of a possible brain injury could potentially be related to all the risk factors listed in . Concentrations of NFL and T-tau gradually increased with time during the first 6 days after trauma. These results are in accordance with previous findings in patients with TBI 
. No concentration peak for GFAP was found, but the boxer with the reported concussion had among the highest concentration levels of GFAP one day after the bout and the levels had decreased from 960 to 500 ng/L at follow up, 15 days after the bout. More studies are needed to investigate if NFL and GFAP correlate with the size of injury.
Our study also revealed higher concentrations of S-100B after a bout compared to controls.
S-100B is a calcium binding protein physiologically produced and released by astrocytes and other glial cells in the central nervous system (CNS) 
. Outside the nervous system it can be found in adipose tissue, muscle and skin 
. S-100B increases after brain injury and remains elevated for up to 5 days in CSF, with a peak at day one 
. The concentrations have been observed to correlate with brain injury severity 
. In serum, S-100B increases after MTBI 
and S-100B levels have also been found to rise after physical activity such as marathon running. S-100B released from skeletal muscle has a relatively short half-life in serum and the levels are back to normal levels within 20 hours 
. To our knowledge no studies have shown transport of S-100B from serum to CSF, why analysis of S-100B in CSF most likely reflects the true cerebral S-100B concentration 
. The role of released S-100B after TBI is not clearly understood but it might have both neurotrophic and neuroprotective functions, or simply reflect injury-related release 
Little is known about the dynamics of Aβ1–42 in CSF but recently the concentrations of Aβ1–42 were demonstrated to correlate with neurological status after acute brain injury. The concentrations were consistent between different sampling occasions in healthy patients but decreased after a brain injury and increased when the neurological status improved 
. Therefore, even though no statistic differences were found, one explanation for the relatively large variation in Aβ1–42 concentrations between test A and B in some of the boxers may be a traumatic brain injury at some stage. In a similar manner the large variation of P-tau with indications of gradual increase in the boxers may be an early sign of neurofibrillary tangle build-up.
The strength of this study is the very well matched baseline parameters for boxers versus controls. The only difference was a longer career of other sports where trauma against head occurs in controls compared to boxers (24% vs 0%) (). Interestingly, the only control subject with elevated levels of NFL and GFAP was among those who had a previous sport career including head trauma. A previous study on CSF biomarkers in soccer players did not find any evidence of TBI in that group 
. A limitation of this study is the variation of time points for CSF sampling. The ideal had been to have test A and B collected at the same time points for all participants. It would also have been preferable with a longer rest period before the B sampling. This was not possible since we had to adapt the samplings to the boxers' schedules in order to perform the study.
In conclusion, this study shows that the repetitive head trauma occurring in olympic boxing may induce changes in CSF NFL, GFAP, T-tau and S-100B, even without anamnestic or clinical symptoms of a concussion or traumatic brain injury. These changes suggest minor central nervous system injuries. It seems that most of the acute injuries can recover with rest but without an appropriate rest period there might be a risk for cumulative injury. The length of the rest period needed seems either to be individual or is correlated to the size of the injury.
Further studies are needed to evaluate if nervous system injury biomarkers in CSF may be useful as an evaluation tool in clinical praxis in the diagnosis and grading of a concussion/MTBI and as part of return to sport guidelines. Future studies of interest include closer monitoring of boxers at different early time-points after repeated head trauma at bouts, long-time follow-up of boxers and also monitoring of CSF biomarkers in patients attending emergency departments due to a concussion where the clinical examination is normal.