Page 33 - JSOM Spring 2025

to the posterior hand (T hand ) (VitalSense). Since warm blood this cutoff imposes a slightly larger penalty on the goodness-
was being removed from the body in a cold environment, it of-fit and will favor a more complex model more frequently. 13
was prudent to examine these thermoregulatory responses. The alpha level in each analysis was set at P=.05, and a Bonfer-
Urine specific gravity was measured to ensure participants be- roni correction was applied to the analyses with multiple com-
gan their blood donation and ruck march hydrated (i.e., urine parisons. Descriptive statistics are presented as means (SDs) or
specific gravity <1.020), and rucksack weights were also mea- medians (interquartile ranges [IQRs]) for ordinal data.
sured during this time.
Results
Immediately after participants completed their BD at the
training area, they were instrumented with heart rate (HR) Ruck Times
monitors (Polar Electro, Bethpage, NY) and portable meta- Results showed a significant order × condition interaction
2
bolic measurement systems (Vyaire Medical, Mettawa, IL) (F (1,11)= 25.4, P<.001, η G=.29), a non-significant main effect of
to measure respiratory rate (RR) and minute ventilation (V ) order (F (1,11) =.22, P>.1, η G=.02), and a non-significant main
2
E
2
during the ruck march. They then donned rucksacks and had effect of condition (F (1,11) =2.13, P>.1, η G=.03). On average,
their oxygen saturation (SpO ) measured using a finger pulse individuals who donated on day 1 were slower to complete the
2
oximeter (Nonin Medical Inc., Plymouth, MN). Participants ruck (68.8 [SD 16.3] min) relative to when they did not donate
9
then provided ratings of perceived exertion (RPE; 1–10 scale), on day 2 (48.6 [SD 8.96] min, P=.004) reflecting a difference
thermal sensation (TS; −4 very cold to +4 very hot), and AMS of 20.2 minutes. Individuals who did not donate on day 1 were
10
11
symptoms. AMS symptoms included headache (0, no head- slower (61.1 [SD 9.0] min) relative to when they did donate
ache to 3, severe, incapacitating headache), fatigue (0, no fa- on day 2 (50.0 [SD 8.9] min), reflecting a difference of 11.1
tigue/weakness to 3, severe fatigue/weakness), and dizziness/ minutes (P=.037). Ruck times did not differ on day 1 across
lightheadedness (0, no dizziness/lightheadedness to 3, severe the order conditions (Welch’s t (8.53) =.85, P>.1), but were overall
dizziness/lightheadedness). Research staff recorded ruck start slower on day 1 (64.0 [SD 15.9] min) relative to day 2 (49.5
and end times with stopwatches. [SD 8.4] min; t =4.49, P<.001).
(12)
Along the 3.2-km route, research staff positioned themselves Physiology
every 0.8 km, creating a series of three checkpoints (CPs) at For heart rate (Figure 3), an LMM (random intercept only)
distances of 0.8km (CP1), 1.6km (CP2), and 2.4km (CP3). At showed a significant effect of HR increase as a function of time
each CP, participants stopped for 2 minutes for the collection (b=4.65 [SE 3.87], P=.009). All other terms were not signifi-
of SpO , RPE, TS, and AMS data. After the 2 minutes elapsed, cant (all Ps>.1). For SpO (random intercept only), an LMM
2
2
participants continued their ruck march to the next CP, where showed a significant effect of SpO decreasing as a function of
2
the procedures were repeated. Six minutes were subtracted from time (b=–1.24 [SE .5], P<.001) and a main effect of condition
participants’ total ruck time to account for stops at each CP. where SpO was lower in the control condition relative to the
2
BD condition (b=–4.23 [SE 2.4], P=.007).
Statistical Analyses
All analyses were completed in R version 4.2.0 (https://www.R- FIGURE 3 Mean (SD) heart rate (HR) responses to ruck marching
project.org/) supported by the rstatix (https://CRAN.R- project. without (CON) and with blood donation (BD) plotted over the
12
org/package=rstatix), lme4 (as described in Bates et al. ), and route. HR increased from ruck marching but was not different
between CON and BD.
ordinal (https://CRAN.R-project.org/package=ordinal) pack-
ages. The total time to complete the ruck march was analyzed
using a 2 (order effect of control/BD: control on Day 1 vs. BD
on Day 1) × 2 (donation condition: control vs. BD) mixed anal-
ysis of variance (ANOVA).
Physiological (HR, SpO , T , T hand, RR, and V ) data were col-
2
c
E
lected as supporting evidence for physical performance out-
comes. These data were recorded at each CP along the route,
including Start and Finish, and analyzed using linear mixed
effects models (LMMs) 2 (control, BD) × 2 (order effect of
control/BD) × 5 (Start, CP1, CP2, CP3, Finish) repeated mea-
sures ANOVA. P values were derived from type II Wald chi-
square tests.
Self-report measures (RPE, TS, headache, fatigue, dizziness/
lightheadedness) were treated as ordinal data and analyzed
using cumulative link mixed models (CLMMs). The model Temperatures
structure was the same as the LMMs. For core temperatures (random intercept only), the LMM
showed a significant effect of core temperatures increasing as
For all models, the construction of random effects structures a function of time (b=.27 [SE .6], P<.001). Furthermore, there
followed a backwards heuristic, starting with the most com- were significant order × time (P=.033) and order × condition
plex random effects structure and working down using likeli- (P=.008) interactions. For the former, core temperatures col-
hood ratio tests (LRT) to attempt to balance model complexity lapsed across conditions were similar and increased across
with goodness of fit (i.e., avoid singular fits and convergence the order groups at times 1, 2, and 3 (Welch’s ts<2.2; Ps>.042
warnings). Thus, we used an α = .2 criteria for LRT, given that corrected).

Performance After Whole Blood Donation at Altitude | 31