Ubiquitin carboxyl-terminal esterase L1 is not elevated in the serum of concussed rugby players: a cross-sectional observational study

The study sample consisted of a total of 79 participants. We recruited 19 people before the game, all women. After the match, 31 individuals (14 men and 17 women) consented to donate blood, among which 5 female individuals also donated their blood before the match. The recreationally active male control group consisted of 12 individuals. Finally, blood sampling for UCHL1 was available for 17 athletes (14 men and 3 women) with CRS. Blood was collected an average of 42.3 min (median=16.0 min, Q1=10.0 min, Q3=45.0 min) after the match in the control group and blood was collected an average of 102, 3 min (median = 37.0 min, Q1 = 10.0, Q3 = 156.0 min) post-concussion in the SRC group.

Demographic comparisons

Comparing demographic data between groups, age was found to be significantly higher in the match-control (median=22.0 years, Q1=19.5 years, Q3=27.5 years, p= 0.010, η2= 0.117) and SRC (median = 24.0 years, Q1 = 20.0 years, Q3 = 26.0 years, p = 0.022, η2= 0.117) compared to the pre-matching group (median = 19 years, Q1 = 18.5 years, Q3 = 20.0 years), but no other differences were found between the groups (p > 0.05) .

When comparing body mass between the groups, it was found that the SRC group (median = 87.3 kg, Q1 = 75.0 kg, Q3 = 90.9 kg) had a significantly higher body mass than the active recreational control group (median = 68.0 kg, Q1 = 63.1 kg, Q3 = 72.9 kg) (p = 0.026, η2= 0.138), but no other difference between groups (p > 0.05, η2= 0.138) were found. Concussion history (number of previous concussions) was significantly different between the matched control group (median=1.0, Q1=0.0, Q3=2.0) and the active recreational control group (median=0.0 , Q1 = 0, Q3 = 0) (p = 0.002, η2= 0.062), with no difference between the rugby groups (p > 0.05). Similarly, years playing rugby were all significantly longer in the rugby groups compared to the non-athlete group (p ≤ 0.002, η2= 0.140), with no difference between the rugby groups (p > 0.05). Table 2 shows the comparison of demographic data between the groups.

Table 2 Descriptive statistics for combined sex data.

UCHL1 Concentration Comparisons

Descriptive statistics of the UCHL1 data used for analysis across all protocols can be viewed in Table 3. There was no significant difference in serum UCHL1 concentration among the four groups (p=0.186, η2= 0.138). UCHL1 concentration was significantly higher in control females (median = 124.3 pg mL−1Q1 = 65.5 pgmL−1Q3 = 261.8 pgmL−1) compared to the pre-match group (median = 17.0 pg mL−1Q1 = 17.0 pgmL−1Q3 = 73.6 pgmL−1) (p = 0.041, η2= 0.080) but there were no differences in men between the groups (p = 0.089, η2= 0.303). Concentration of UCHL1 in corresponding female control (median = 124.3 pg mL−1Q1 = 65.5 pgmL−1Q3 = 261.8 pgmL−1) was significantly higher than the concentration of UCHL1 in the corresponding control male (median = 8.5 pg mL−1Q1 = 0.0 pgmL−1Q3 = 17.0 pgmL−1) (p = 0.002, d = 0.590), but there was no difference between the recreational active control group (median = 263.7 pg mL−1Q1 = 0.0 pgmL−1Q3 = 822.1 pgmL−1) and pre-match group (median = 17.0 pg mL−1 Q1 = 17.0 pgmL−1Q3 = 73.6 pgmL−1) (p = 0.236, d = 1.148) or between women (median = 17.0 pg mL−1Q1 = 8.5 pgmL−1Q3 = 161.3 pgmL−1) and male (median = 17.0 pg mL−1Q1 = 0.0 pgmL−1Q3 = 17.0 pgmL−1) UCHL1 concentration in SRC group (p=0.591, d=0.752). Descriptive statistics comparing the concentration of UCHL1 in sex-matched data and between male and female data can be viewed in Table 3.

Table 3 Serum concentration of ubiquitin carboxyl-terminal esterase L1 (UCHL1) between groups and between sexes.

Diagnostic accuracy

The AUROC of UCHL1 comparing all groups can be found in Table 4. All comparisons show that UCHL1 did not discriminate between groups in this study (AUROC 0.05). We also compared sex-matched groups (Table 5) and found significance between the pre-matching group and the female matching control group (AUROC = 0.745, p = 0.012). However, comparison of all other sex-matched groups showed that UCHL1 did not discriminate between groups (AUROC 0.05).

Table 4 Area under the receiver operating characteristic curve (AUROC) comparing different groups using serum.
Table 5 Area under the receiver operating characteristic curve (AUROC) comparing sex-matched groups using serum.

Sensitivity analysis

Since age has been found to be a confounding variable in other brain biomarkers22we removed nine subjects over the age of 40 and rerun the statistics. Removal of this data did not affect the statistical analysis, so these subjects were left in the final data pool. A sensitivity analysis was also conducted on the timing of blood collection in the match-control and SRC groups, resulting in the removal of one subject from the match-control group and two subjects from the SRC group. Removing these data points did not change the statistical result, so these subjects were left in the final data pool. Sensitivity analysis conducted on serum UCHL1 concentrations found one data point in the matched female control group (UCHL1 concentration = 2805.5 pg mL−1) as an outlier. After further analysis of this subject’s demographics and rugby match ratings, we found no logical reason to disregard this subject’s data in the analysis. However, it is important to note that when this data point was removed from the analysis, the statistical difference between the pre-match group and the female match control group is no longer statistically significant with a p-value of 0.069 vs. 0.041 when the data point is left in the dataset. However, removing this data point did not change the lack of significance among the combined sex groups nor did it affect the significance, or lack thereof, for the AUROC of UCHL1.