Page 98 - JSOM Spring 2020

more upbeat positional nystagmus immediately after the blasts dichotomous in some, objectively assessed using gauges in oth-
than during baseline, although this was not statistically signif- ers), different outcome measures examined, different inclusion
icant. Because only horizontal smooth pursuit was assessed in and exclusion criteria (particularly with regard to history of
the study, the authors noted the importance of studying ver- TBI and blast history exposure), and different statistical ana-
tical smooth pursuit in future studies. There is no consistent lytic strategies. Although some studies corrected for multiple
finding suggestive of vestibular dysfunction associated with statistical comparisons, most did not, increasing the likelihood
repeated low-level blast exposure. of Type I errors (i.e., false positives). Also, some studies in-
cluded control groups and others did not. For those relying on
Vision within-subject changes over time, it was often unclear if only a
In the only study to specifically address vision, Capo-Aponte priori comparisons were made rather than making every pos-
et al. compared blast-exposed breacher instructors with non- sible comparison on any number of variables. Although this is
28
exposed breacher engineers 10 times over the course of 2 years understandable in such a new area of inquiry, it does decrease
and found no differences in visual acuity, contrast sensitivity, confidence in findings, given the small number of studies and
oculomotor function, color vision, pupillary light reflex, cor- small number of participants. Very few studies reported all sta-
neal thickness, or retinal/optic nerve integrity. However, the tistics associated with (null) findings or effect sizes.
instructors had a decrease in visual field sensitivity and corneal
endothelial cell density. The authors noted that near-vertical The sample sizes in these studies were universally small. This is
phoria (NVP) was the only tested oculomotor function that not surprising, given the small size of breacher training courses
appeared to be affected by the level of blast exposure during and the recent interest in this topic. In addition to low power
the study. However, the NVP values were within normal lim- to detect any potential effects, small sample sizes also illus-
its, and the differences between groups were not statistically trate the importance of handling outliers. For example, in the
significant. Carr et al. study, there was an individual who had a greater
13
than a 1,000% increase from baseline level in the blood bio-
marker being studied (i.e., UCH-L1), but the magnitude of his
Discussion
blast exposure was less than 2 psi. In contrast, another indi-
We reviewed 18 studies of low-level blast exposure, as experi- vidual had blast exposure greater than 5 psi (high exposure in
enced during military and police training. Although there were this particular study), yet a blood sample showed a decline in
inconsistencies in findings between studies, some trends were UCH-L1 level from baseline. The authors noted that the time
apparent. For blood-based biomarkers, it may be important to between blast exposure and blood sample collection could not
look at peak pressure, as opposed to total exposure, to find be controlled in this study, which may have contributed to the
13
an effect, because biomarker changes may be ephemeral and heterogeneity in findings between individuals.
seen only at peak individual times. The extent to which eleva-
tion of biomarker levels may be due to peripheral nervous sys- All reviewed studies had high risk for bias due to some com-
tem injury is unclear; correlation with neuroimaging findings bination of nonblinding, selection bias, selective outcome
is needed. Epigenetic changes were noted by Gill et al. after reporting, lack of measurement of exposure (or measuring
18
moderate blast exposure, with the APP network being most exposure but not using those data), and not controlling for
affected. The changes in APP were detected up to 2 days later potential confounds (e.g., age, history of TBI). Furthermore,
after moderate blast exposure, but then returned to baseline from a methodological standpoint, very few studies used reli-
(3 days after). Generally, there was no evidence of consistent able change indices (RCI) when examining changes over time.
neurocognitive effects of blast exposure, though it may be that RCIs help determine what change is needed on a measure to
typical measures used are not sensitive enough, as has been be reliable and not due to chance or measurement error, taking
illustrated in the concussion literature. Two studies 16,22 re- into account that measure’s distribution of change scores and
30
ported an acute or postacute effect on a few tests within the including test-retest reliability. 31
context of many comparisons. Neuroimaging studies suggest
some acute to postacute increases in activation during a work- Perhaps the most concerning issue is the lack of attempts to
ing-memory task associated with peak overpressure, though correlate outcome measures with objective blast-exposure
there is no evidence that these changes endure. data. This is likely due to the difficulty of extrapolating data
from blast-gauge data, particularly over time. Using blast
The preponderance of evidence suggests no lasting effect of gauges provides an objective measure of exposure; however,
blast exposure on symptom complaints. Increased headaches it is difficult to know how individual pressure-sensor readings
may be experienced in a subset of individuals, but only acutely. represent reflective versus incident pressure. The orientation
There was no consistent finding suggestive of vestibular dys- of the wearer affects these variables, as does the environment
function associated with repeated low-level blast exposure. Fu- (e.g., reflective surfaces, use of multiple charges). Further-
ture studies should further investigate possible isolated effects more, blast gauges produce a plethora of data, including im-
among those with positive TBI histories. Only one study spe- pulse pressure, peak pressure, interval between peaks, changes
28
cifically assessed vision multiple times across 2 years; no dif- in these measures, and so forth. Automated algorithms are
ferences were found between breachers and control subjects. needed to provide standardized blast metrics that are indepen-
dent of orientation and other confounds. Within the US Army
Considerable heterogeneity among the studies makes drawing Special Operations Command, a program called CONQUER
any firm conclusions difficult, including different evaluation will surveil Operators using blast gauges in an ongoing effort
periodicity, different levels of blast exposure for included par- to further characterize blast exposure.
ticipants (in terms of strength and duration), different levels
of protection offered the participants during training (often Findings of differences between instructors and students, de-
not detailed), different ways of assessing “exposure” (i.e., spite the latter having higher blast exposure during the acute

92 | JSOM Volume 20, Edition 1 / Spring 2020