Hindawi

Depression and Anxiety

Volume 2023, Article ID 6869543, 11 pages
https://doi.org/10.1155/2023/6869543

Research Article
The Effect of a Combined Mindfulness and Yoga Intervention on
Soldier Mental Health in Basic Combat Training: A Cluster
Randomized Controlled Trial

Thomas H. Nassif ,1 Ian A. Gutierrez ,1 Carl D. Smith ,1 Amishi P. Jha ,2

and Amy B. Adler 1
1
Walter Reed Army Institute of Research, 503 Robert Grant Avenue, Silver Spring, MD 20910, USA
2
University of Miami, Coral Gables, FL 33124, USA

Correspondence should be addressed to Thomas H. Nassif; [email protected]

Received 18 October 2022; Revised 11 August 2023; Accepted 24 October 2023; Published 1 December 2023

Academic Editor: Carlos Carona

Copyright © 2023 Thomas H. Nassif et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.

Background. Depression, anxiety, and sleep problems are prevalent in high-stress occupations including military service. While
effective therapies are available, scalable preventive mental health care interventions are needed. This study examined the
impact of a combined mindfulness and yoga intervention on the mental health of soldiers in Basic Combat Training (BCT).
Methods. U.S. Army soldiers (N = 1,896) were randomized by platoon to an intervention or training-as-usual condition.
Soldiers in the intervention condition completed Mindfulness-Based Attention Training (MBAT), engaged in daily 15 min
mindfulness practice, and participated in 30 minutes of hatha yoga 6 days per week. Surveys were administered at baseline
(T1, prior to training), week 4 of BCT (T2), week 6 (T3), and week 9 (T4). Results. A significant time-by-condition interaction
predicting positive screens for depression found that screens decreased at a faster rate from T1 to T4 in the intervention
condition (-12.6%) compared to training-as-usual (-7.2%) (b = −0 18, SE = 0 07, p = 0 028). While positive anxiety screens
decreased over time across conditions, the time-by-condition interaction found no significant differences in the rate of these
decreases by condition (b = 0 09, SE = 0 09, p = 0 273). A significant time-by-condition interaction predicting positive screens
for sleep problems found that sleep problems decreased in the intervention condition (-1.4%) but increased in training-as-
usual (2.0%) (b = −0 68, SE = 0 16, p = 0 027). Conclusion. The mindfulness and yoga intervention was associated with a greater
reduction in positive screens for depression and sleep problems among soldiers during high-stress training. Limitations include
reliance on self-report and the inability to disaggregate the effects of mindfulness versus yoga. Mindfulness and yoga may
enable personnel in high-stress occupations to sustain their mental health even in the context of significant psychological
demands. This trial is registered with NCT05550610.

1. Introduction While effective therapies are available in the military

health system [7], the demands on health care resources
Depression, anxiety, and sleep problems are prevalent in high- [8, 9] signal the need for scalable, upstream interventions
stress occupations including health care [1, 2], first responding that can be implemented early in a soldier’s career. To
[3, 4], and military service [5]. These mental health concerns address this need for preventive mental health care strate-
are associated with missed work days and impaired workplace gies and early intervention in the military [10], the present
functioning [6]. Given the prevalence and implications of study examined the impact of a combined mindfulness and
these mental health problems in high-stress occupations like yoga intervention on mental health among soldiers in Basic
military service, interventions are urgently needed. Combat Training (BCT).
2 Depression and Anxiety

The U.S. Army chose to evaluate a combined mindful- These findings have been extended to high-stress work
ness and yoga intervention because they wanted to leverage environments. For example, studies conducted in the
the benefits of both interventions while taking into account Indian military have found that yoga was associated with
funding constraints, timeline considerations, and platoon decreases in anxiety symptoms and improvements in sleep
availability. These two interventions provide complementary quality [29, 30]; however, in these studies, yoga was pro-
skill sets that offer a potentially robust foundation for occu- vided during an intensive residential program that included
pational stress tolerance and primary mental health preven- other components (e.g., guided relaxation, prayer, chanting,
tion among service members. Consistent with this and philosophy [29]). This larger, intensive program dem-
integration of the two interventions, some commonly onstrates the need to assess the feasibility of scaling yoga
researched mindfulness programs include postural yoga for a broader military context.
components [11, 12], and many postural yoga programs The present study reports on a cluster randomized con-
incorporate elements of mindfulness [13, 14]. trolled trial evaluating a combined mindfulness training and
Mindfulness is a mental mode involving focused attention postural yoga intervention piloted by the U.S. Army during
to the present-moment without elaboration or emotional reac- BCT. Incoming service members are less likely to have pre-
tivity [12, 15]. Mindfulness training programs can enhance conceived notions of current Army training and receive
concentrative attention, cognitive monitoring, and openness instructions in a highly structured environment. Thus,
to experience to help decrease mind-wandering and manage BCT was selected to enable the implementation of a novel
stress [16]. Ample evidence supports the effectiveness of training method that could help new soldiers manage occu-
mindfulness training for addressing mental health in clinical pational stress and prevent mental health problems. Ran-
contexts [17, 18]. Mindfulness training has also been found domization occurred at the platoon level to be consistent
to reduce anxiety and depression among personnel across a with military training delivered within the BCT context.
range of job sectors [19]. In addition, mindfulness training To our knowledge, this is the largest randomized trial to
has been found to improve sleep quality and duration, alleviate examine the effectiveness of mindfulness and yoga interven-
insomnia symptoms, and mitigate cognitive processes that tions on mental health outcomes in the military context.
interfere with sleep, such as worry and rumination among Given the prevalence of depression, anxiety, and sleep prob-
civilians and veterans [20, 21]. lems and the lack of scalable, preventive mental health care
Although some studies with military personnel have not strategies in the military, findings from this study may
found an impact of mindfulness training on anxiety and inform primary mental health prevention among service
depression symptoms [22, 23], studies have found that members. In addition, this study addresses the literature
greater frequency of mindfulness practice has led to gaps on mindfulness practices that are embedded into work
improved emotional health among service members, such activities. Such embedded practices can inform how formal
as positive mood [15] and emotion regulation [22]. Mindful- training programs can be implemented into an occupational
ness intervention studies have operationalized formal mind- setting. Furthermore, there is limited research on the effects
fulness practice as a structured session in which participants of a scalable yoga intervention conducted with military pop-
follow a prerecorded mindfulness exercise. While formal ulations. More broadly, no previous research has examined
practice is beneficial [24], it may be important to supplement the impact of mindfulness training and yoga among a large
training programs with embedded practices that are incor- cohort of basic combat trainees.
porated into work activities. Although little is known about Our primary aim was to determine the effect of a mind-
the effects of embedding practice, this research gap is essen- fulness and yoga intervention on mental health outcomes.
tial to address given the potential advantage of bringing We predicted that soldiers receiving the mindfulness and
mindfulness into work-related activities. yoga intervention would be less likely to screen positive for
Another intervention that may be useful in supporting depression, anxiety, and sleep problems over time compared
service member mental health is postural yoga. Postural to the training-as-usual condition. Our secondary aim was
yoga emphasizes physical postures and breathing tech- to determine the effectiveness of embedded mindfulness
niques that have been used to support mental well-being practice on these outcomes; we predicted that more frequent
[25]. Although yoga intervention studies predominantly embedded mindfulness practice would lead to fewer positive
employ physical postures as the core component of the screens. Although this study design did not allow for distin-
study intervention, other practices (e.g., meditation [14]) guishing between the unique effects of mindfulness training
are occasionally included as part of a multicomponent and yoga, the overall purpose of this study was to inform
yoga intervention. health promotion efforts regarding the potential benefits of
While there is evidence that postural yoga interventions implementing a combined mindfulness and yoga interven-
improve depression symptoms in clinical samples [26], the tion during BCT and to guide the development of future
effectiveness of yoga interventions for anxiety disorders intervention programming.
and sleep problems is limited [13]. Still, yoga interventions
have been found to decrease time to fall asleep and increase 2. Materials and Methods
the number of hours slept [27]. With respect to the occupa-
tional context, yoga interventions in workplace settings are 2.1. Participants. The study was designed as a cluster
associated with improvements in anxiety symptoms and randomized trial. Participants were U.S. Army soldiers
sleep quality, but not changes in depression [28]. (N = 1,896) from two battalions who attended BCT between
Depression and Anxiety 3

Table 1: Demographics by condition.

Condition
Total
Variable Mindfulness and yoga Training-as-usual Test statistic p
N (percent) N (percent) N (percent)
N 1584 813 771
Age χ2 3 = 0 32 0.956
18-19 664 (42.8%) 341 (43.2%) 323 (42.4%)
20-24 525 (33.8%) 262 (33.2%) 263 (34.5%)
25-29 221 (14.2%) 114 (14.4%) 107 (14.0%)
30+ 142 (9.1%) 73 (9.2%) 69 (9.1%)
Education χ2 3 = 6 42 0.093
HS/GED 908 (57.6%) 462 (57.1%) 446 (58.1%)
Some college or associates 449 (28.5%) 220 (27.2%) 229 (29.8%)
Bachelors 175 (11.1%) 105 (13.0%) 70 (9.1%)
Graduate 45 (2.9%) 22 (2.7%) 23 (3.0%)
Gender χ2 1 ≤ 0 01 >0.999
Male 1142 (72.2%) 586 (72.3%) 556 (72.2%)
Female 439 (27.8%) 225 (27.7%) 214 (27.8%)
Past yoga practice 777 (49.8%) 409 (50.9%) 368 (48.6%) χ2 1 = 0 75 0.387
2
Past mindfulness practice 749 (48.1%) 383 (47.8%) 366 (48.4%) χ 1 = 0 04 0.835
Baseline mental health
Positive depression screens 589 (37.9%) 310 (38.8%) 279 (37.0%) χ2 1 = 0 46 0.498
2
Positive anxiety screens 825 (53.2%) 405 (50.9%) 420 (55.6%) χ 1 = 3 32 0.068
2
Positive sleep problem screens 94 (6.0%) 61 (7.6%) 33 (4.4%) χ 1 = 6 74 0.009∗∗
Note. Percentages reported as a total of valid responses. ∗ p < 0 05, ∗∗ p < 0 01, and ∗∗∗
p < 0 001.

October 2020 and December 2020. BCT introduces enlisted 2.2. Intervention. Mindfulness-Based Attention Training
soldiers to the Army by teaching them foundational soldier (MBAT) is an 8-hour manualized mindfulness intervention
skills. The structured daily training schedule ranges from developed and contextualized for delivery to military
12 to 15 hours per day for ten weeks. Soldiers live together personnel [16, 23, 31]. MBAT was delivered in a classroom
in barracks and work together in platoons. During BCT, sol- setting or comparable low-distraction environment. The
diers are required to become proficient in a series of tasks course consists of 2-hour sessions each week for 4 weeks,
including hand-to-hand combat, rifle marksmanship, com- reflecting four themes: (1) breath awareness and focused
bat lifesaver skills, and land navigation. This training occurs attention skills, (2) body awareness without judgment, (3)
in classrooms, ranges, and field environments. open monitoring to observe sensory and mental experiences,
In BCT, each battalion consisted of five companies, and and (4) interpersonal connection. Each session introduced a
each company consisted of four platoons. In the Basic Com- 15-minute mindfulness practice: (1) focused attention, (2)
bat Training context, soldiers are arbitrarily assigned to pla- mindful body scan, (3) open monitoring, and (4) connec-
toons, companies, and battalions without regard to tion. Group mindfulness practice 6 days per week consisted
individual characteristics. Since the Army randomly assigns of listening to a 15-minute audio recording. Each session
soldiers to platoons, we randomized platoons one and two also presented an exercise for embedding mindfulness prac-
to mindfulness training and yoga and platoons three and tice during the day.
four to training-as-usual. This approach facilitated the mili- MBAT was delivered by masters- or doctoral-level per-
tary and research team in tracking a large number of units as formance experts (n = 10) who had an average of 4.40 years
part of implementing an evaluation in the real world. In all, (SD = 3 62) of experience working within the military and
20 platoons were assigned to mindfulness and yoga (n = 813), little to no prior experience with mindfulness. Instructors
and 20 platoons were assigned to training-as-usual (n = 771). attended 26 hours of virtual training sessions, participated
Soldiers in both conditions participated in the evaluation in five hours of small group activities over the course of 12
throughout the ten weeks of BCT. Participant demographics weeks, and were assigned 15 minutes of daily mindfulness
reflected a typical BCT population. Table 1 provides demo- practice. Two instructors were assigned to each intervention
graphic details for the sample along with comparisons by platoon; each platoon was divided in half to ensure a smaller
condition. instructor-to-soldier ratio.
4 Depression and Anxiety

The postural yoga program consisted of hatha yoga pos- to harm self or others. To minimize coercion or undue influ-
tures that replaced exercises normally included at the begin- ence, drill sergeants and other unit leadership were not pres-
ning (preparatory drills) and end (recovery drills) of daily ent during the consent process. Study procedures were
Physical Readiness Training (PRT). Yoga postures were conducted in accordance with Army Regulation 70-25 [34].
designed to engage major muscle groups during the prepara- As an exempt study, surveys and procedures were approved
tory drills (e.g., sun salutations, crescent lunge, and eagle pose) by the Human Subjects Protection Branch at the Walter
and to release tension during the recovery drills (e.g., gate Reed Army Institute of Research.
pose, reverse plank, and bridge pose). Online resources pro-
vide additional information about these yoga postures [32, 33]. 2.4. Inclusion Criteria. Soldiers attending BCT at one mili-
Yoga instructors (n = 10) had at least two years of teach- tary installation in the Southeastern United States were eligi-
ing experience and were registered with Yoga Alliance at the ble for inclusion.
Registered Yoga Teacher (RYT) 200-hour level or higher. 2.5. Exclusion Criteria. Soldiers who tested positive for
One yoga instructor was assigned to each intervention pla- COVID-19 during routine BCT medical screening were
toon. Yoga instructors completed 8 hours of virtual training excluded from analysis.
sessions, which focused on teaching a standardized sequence
of yoga postures aligned with the physical demands of BCT,
3. Measures
linking movement to the breath, stabilizing core muscles,
building lower and upper body strength, and reducing ten- 3.1. Demographics. At T1, soldiers were asked demographic
sion in the low back and hips. questions. They were asked their current age with
response options across four categories (18-19, 20-24,
2.3. Procedure. Soldiers in the intervention condition com- 25-29, and 30+). In terms of gender, they were asked if
pleted the MBAT course during the first four weeks of they were male or female. They were also asked their
BCT, engaged in daily 15 min mindfulness practice with highest level of civilian education with response options
their platoon, were instructed on embedding individual across four categories (high school or equivalent, some
mindfulness practice into the duty day, and participated in college or associate degree, bachelor’s degree, and gradu-
30 minutes of hatha yoga 6 days per week. While platoons ate degree).
in the intervention condition were participating in interven-
tion activities, platoons assigned to training-as-usual either 3.2. Mindfulness and Yoga History. At T1, soldiers were
were instructed to use the time to review BCT material or asked how many times they had practiced mindfulness or
engaged in standard preparatory and recovery drills. The yoga before BCT using responses ranging from 0 (never) to
standard preparatory drills consisted of exercises (e.g., rear 3 (many times). The mindfulness item and the yoga item
lunge, high jumper, and squat bender) designed to increase were dichotomized into “no prior experience” and “prior
muscular endurance, reduce injury risk, and improve overall experience,” with responses greater than “never” being
fitness. The standard recovery drills consisted of stretches (e. coded as “prior experience.”
g., overhead arm pull, extend and flex, and thigh stretch) to
develop flexibility across major muscle groups including the 3.3. Depression. Depression was assessed with the two-item
shoulders, hip flexors, and lower back. Patient Health Questionnaire (PHQ-2) [35] rated from 0
All soldiers were briefed prior to study enrollment, and (not at all) to 3 (nearly every day). The two items were
soldiers who provided informed consent for research pur- summed, and scores of 3 or greater were coded as positive
poses did so prior to completing the baseline survey. Soldiers screens. Scores were not tabulated if data were missing for
were first assigned to a residential bay upon arrival at BCT; either item. The PHQ-2 has been shown to be a reliable
after three weeks, bays were then allocated to larger pla- and valid measure in previous studies [35–37].
toons. Bays were randomized to either an intervention or 3.4. Anxiety. Anxiety was measured with the two-item Gen-
training-as-usual platoon. Participant exclusion, randomiza- eralized Anxiety Disorder (GAD-2) scale [38] rated from 0
tion, and attrition numbers are presented in the CONSORT (not at all) to 3 (nearly every day). The two items were
diagram in Figure 1. Of the 1,896 soldiers in the evaluation, summed, and scores of 3 or greater were coded as positive
data from 312 (16.5%) were excluded from analysis: 81 did screens. Scores were not tabulated if data were missing for
not consent to have their survey responses used for research either item. The GAD-2 has been shown to be a reliable
purposes, 177 could not be matched to condition assignment and valid measure in previous studies [36, 39].
by their unique identifier, and 54 were assigned to a bay of
COVID-19 positive soldiers. 3.5. Sleep Problems. Sleep problems were measured using
There were four surveys: baseline (T1, prior to training), four items from the Insomnia Severity Index [40]; this short-
week 4 of BCT (T2), week 6 (T3), and week 9 (T4). Soldiers ened form has been used in other military studies [41]. Sol-
in the intervention condition also completed a course evalu- diers rated their sleep over the past two weeks with respect to
ation at T4. Completion of surveys was voluntary. Soldiers their difficulty falling asleep and difficulty staying asleep
were instructed that there were no penalties for choosing from 1 (none) to 5 (very severe), their satisfaction with their
not to participate and they would not be compensated for current sleep pattern from 1 (very dissatisfied) to 5 (very sat-
participation. Soldiers were informed that their responses isfied), and the extent to which their sleep interfered with
were private and confidential unless they indicated intent their daily functioning from 1 (not at all/no sleep problem)
Depression and Anxiety 5

Enrollment Assessed for eligibility

(kplatoons= 40, n = 1,896)

Excluded (n = 312)
(i) Did not provide consent for data to be
used for research purposes (n = 81)
(ii) Unable to match data (n = 177)
(iii) Positive for COVID-19 (n = 54)

Randomized
(k = 40, n = 1,584)

Allocated to MT and Yoga Intervention (n = 813) Allocated to Training-as-Usual Condition (n = 771)

Allocation and
(i) Received allocated intervention (k = 20, n = 813) (i) Received allocated intervention (k = 20, n = 771)
Baseline (T1)
(ii) Did not receive allocated intervention (k = 0, n = 0) (ii) Did not receive allocated intervention (k = 0, n = 0)

Training
Received Training
Administration

Follow-up (T2) Lost to follow-up/unmatched data (k = 0, n = 0) Lost to follow-up/unmatched data (k = 0, n = 0)

Follow-up (T3) Lost to follow-up/unmatched data (k = 0, n = 44) Lost to follow-up/unmatched data (k = 0, n = 42)

Follow-up (T4) Lost to follow-up/unmatched data (k = 0, n = 122) Lost to follow-up/unmatched data (k = 0, n = 129)

Completed All Phases Completed All Phases (k = 20, n = 691) Completed All Phases (k = 20, n = 642)

Figure 1: CONSORT diagram.

to 5 (very much). Positive screens were based on scoring “at likelihood of positive screens by time and condition while
risk” on at least 3 items [41]. Scores were not tabulated if accounting for potential clustering effects of platoon mem-
data were missing for any item. bership. GLMMs were estimated using restricted maximum
likelihood estimation with bound optimization by quadratic
3.6. Individual Mindfulness Practice during BCT. On the approximation. Each GLMM was specified as a three-level
course evaluation survey, soldiers were asked, “On average, longitudinal model, with time nested within soldier and sol-
how many days a week did you practice MBAT on your diers nested within platoons. Random intercepts and uncor-
own?” Response options were “0 days,” “1-2 days,” “3-4 related random slopes were estimated for all models.
days,” “5-6 days,” and “7 days.” Soldiers who reported prac- Interclass correlation coefficients (ICCs) were computed to
ticing 1-2 days per week or less were classified as “low prac- assess the proportion of variance explained by the random
tice,” while those who reported practicing 3-4 days per week effects of soldier and platoon.
or more were classified as “high practice.” For each outcome, positive mental health screens were
first modeled as a function of the interaction between time
4. Statistical Analysis and condition, with training-as-usual defined as the referent
group (hereafter referred to as “time-by-condition”
Frequencies were calculated for sample demographics, and GLMMs). Subsequently, to examine how the frequency of
chi-square tests were conducted to test demographic differ- individual embedded mindfulness practice impacted positive
ences between conditions to identify potential covariates mental health screens, we then ran another set of GLMM
for subsequent models. modeling positive screens as a function of an interaction
The percentage of soldiers screening positive for depres- between time and condition, with the mindfulness and yoga
sion, anxiety, and sleep problems by condition was com- condition divided into the low practice and high practice
puted for each time point. We employed binomial groups (hereafter referred as “time-by-practice” GLMMs).
generalized linear mixed models (GLMMs) to predict the Across all six GLMMs—three time-by-condition models
6 Depression and Anxiety

and three time-by-practice models—training-as-usual was 5.2. Anxiety. ICCs for anxiety were 0.47 for the random
treated as the reference group. effect of soldier and less than 0.01 for the random effect of
Across all conditions and time points, items assessing platoon. The time-by-condition model identified a signifi-
depression, anxiety, and sleep problems had no more than cant main effect of time on positive screens, such that anxi-
3.6% missing data. Missing data for positive screens were ety decreased across the entire study sample (Table 2).
4.1% for depression, 4.1% for anxiety, and 2.7% for sleep Neither a significant effect of condition nor a significant
problems. The individual mindfulness practice item had time-by-condition was observed (see Figure S1B).
5.2% missing data. For chi-square analyses conducted on The time-by-practice model also did not yield significant
demographic data, listwise deletion was used to remove group- or condition-wise differences. The proportion of sol-
respondents who did not provide information. For the diers screening positive among the high practice group, low
time-by-condition and time-by-practice models, missing practice group, and training-as-usual condition decreased
data were handled through intent-to-treat analyses through 31.7%, 30.5%, and 32.6% from T1 to T4, respectively
the use of GLMMs [42]. (Figure 2(b); confidence interval bands are provided in
We conducted an a priori multilevel power analysis Figure S2B).
using Optimal Design software [43]. Analyses found that
power exceeded 0.8 for an estimated effect size of d = 0 18 5.3. Sleep Problems. ICCs for sleep problems were 0.91 for
with cluster ICCs not exceeding 0.01 (k = 40 platoons and, the random effect of soldier and less than 0.01 for the ran-
approximately, n = 40 soldiers per platoon). This result indi- dom effect of platoon. The time-by-condition model
cated that our design would be adequate to identify small to revealed a significant main effect of time, indicating a signif-
medium effects. icant decrease in positive screens across conditions. We also
All analyses were conducted in R v.4.1.0 [44]. GLMMs observed a significant main effect of condition, highlighting
were estimated using the “lme4” package [45]. the overall higher level of sleep problems among the inter-
vention condition, which is most evident at T1 where sleep
5. Results problems were 3.4% higher in the intervention condition
compared to training-as-usual. A significant time-by-
No significant condition-wise differences were observed with condition interaction reflected differences in condition-wise
respect to demographics and mindfulness and yoga history slopes, such that sleep problems in the intervention condi-
(Table 1). Thus, these variables were excluded from further tion decreased 1.4% from T1 to T4, whereas sleep problems
analyses. Among soldiers in the intervention condition in training-as-usual increased 2.0% (see Figure S1C).
who indicated their frequency of individual embedded Figure 2(c) depicts the time-by-practice model for sleep
mindfulness practice (83.0%, n = 677), 37.0% (n = 249) problems, demonstrating a significant interaction such that
reported high individual practice. the proportion of soldiers with sleep problems in the high
practice group decreased, whereas sleep problems among
5.1. Depression. ICCs for depression were 0.53 for the ran- soldiers in training-as-usual increased (confidence interval
dom effect of soldier and less than 0.01 for the random effect bands are provided in Figure S2C). This model found no
of platoon, indicating that depression outcomes shared significant difference in sleep problems over time between
intraindividual variability across time but evidenced little the low practice group and training-as-usual, indicating
clustering by platoon. The time-by-condition model found that the significant interaction observed in the time-by-
a significant main effect of time, such that positive depres- condition model is better understood as a being driven by
sion screens decreased over time across all soldiers the high practice group. From T1 to T4, sleep problems in
(Table 2). However, the model also revealed a significant the high practice group fell 3.8%, whereas positive screens
interaction between time and condition, such that the pro- in the low practice group fell 0.7%.
portion of positive screens among soldiers in the mindful-
ness and yoga condition decreased at a significantly faster 6. Discussion
rate than did the proportion of positive screens among sol-
diers in the training-as-usual condition: from T1 to T4, pos- Soldiers assigned to a combined mindfulness and yoga train-
itive screens fell 12.6% in the mindfulness and yoga ing during BCT were less likely to have positive screens for
condition, whereas positive screens fell 7.2% in the depression or sleep problems over time compared to those
training-as-usual condition (see Figure S1A). receiving training-as-usual. Despite baseline differences in
The time-by-practice model revealed a significant time- sleep problems, analyses examining a time-by-condition
by-practice interaction, such that the proportion of soldiers interaction, which accounts for baseline differences, revealed
screening positive in the high practice group decreased at a a significant effect for the intervention condition. Findings
faster rate compared to soldiers in training-as-usual were consistent with prior research demonstrating the salu-
(Table 2). There was no significant difference in positive tary effects of mindfulness training on depression [17, 18, 46,
screens over time between the low practice group and 47] and insomnia symptoms [20, 21] and of postural yoga
training-as-usual. From T1 to T4, positive screens in the on depression symptoms [26] and sleep quality [27].
high practice group fell 21.8%, whereas positive screens in We also found that positive screens for depression and
the low practice group fell 7.3% (Figure 2(a); confidence sleep problems decreased more rapidly over time among sol-
interval bands are provided in Figure S2A). diers who engaged in individual embedded mindfulness
 Table 2: Generalized linear mixed effect models predicting mental health screens from time-by-condition and time-by-practice interactions.

Depression Anxiety Sleep problems

Depression and Anxiety

b SE p OR [95% CI] b SE p OR [95% CI] b SE p OR [95% CI]

Time-by-condition models
Fixed effects
Intercept -0.32∗ 0.11 0.043 0.73 [0.54, 0.99] 1.27∗∗∗ 0.57 <0.001 3.56 [2.60, 4.88] -7.27∗∗∗ 0.00 <0.001 0.00 [0.00, 0.00]
∗∗ ∗∗∗
Time -0.26 0.06 0.001 0.77 [0.67, 0.89] -0.92 0.03 <0.001 0.40 [0.34, 0.46] -0.78∗∗ 0.11 0.001 0.46 [0.29, 0.74]
Condition1 0.28 0.25 0.145 1.32 [0.91, 1.92] -0.29 0.15 0.155 0.75 [0.50, 1.12] 1.44∗ 2.60 0.018 4.24 [1.28, 14.08]
1
Time × condition -0.18∗ 0.07 0.028 0.83 [0.71, 0.98] 0.09 0.09 0.273 1.10 [0.93, 1.29] -0.68∗ 0.16 0.027 0.51 [0.28, 0.93]
Random effects Variance Variance Variance
Soldier-intercept 1.62 1.92 6.58
Soldier-slope 0.54 0.43 2.81
Platoon-intercept <0.01 0.00 <0.01
Platoon-slope 0.10 0.09 <0.01
Time-by-practice models b SE p OR [95% CI] b SE p p b SE p OR [95% CI]
Fixed effects
Intercept -0.36∗ 0.11 0.021 0.70 [0.52, 0.95] 1.26∗∗∗ 0.57 <0.001 3.53 [2.58, 4.84] -7.35∗∗∗ 0.00 <0.001 0.00 [0.00, 0.00]
∗∗ ∗∗∗ ∗∗
Time -0.24 0.05 0.001 0.79 [0.69, 0.90] -0.92 0.03 <0.001 0.40 [0.34, 0.47] -0.75 0.11 0.002 0.47 [0.29, 0.76]
Low practice 0.12 0.26 0.592 1.13 [0.72, 1.78] -0.26 0.19 0.275 0.77 [0.48, 1.23] 1.32 2.74 0.071 3.74 [0.89, 15.68]
High practice 0.55 0.49 0.053 1.73 [0.99, 3.02] -0.15 0.25 0.598 0.86 [0.48, 1.52] 2.31∗∗ 8.29 0.005 10.08 [2.01, 50.49]
Time × low practice1 -0.07 0.09 0.441 0.93 [0.78, 1.12] 0.08 0.10 0.382 1.09 [0.90, 1.31] -0.65 0.20 0.083 0.52 [0.25, 1.09]
1
Time × high practice -0.45∗∗∗ 0.07 <0.001 0.64 [0.51, 0.80] -0.01 0.11 0.931 0.99 [0.79, 1.24] -1.21∗ 0.15 0.013 0.30 [0.12, 0.77]
Random effects Variance Variance Variance
Soldier-intercept 1.72 1.90 6.70
Soldier-slope <0.01 <0.01 <0.01
Platoon-intercept 0.47 0.43 2.79
Platoon-slope 0.09 0.10 <0.01
Note. 1Training-as-usual used as the referent group. ∗ p < 0 05, ∗∗ p < 0 01, and ∗∗∗ p < 0 001.
7
8 Depression and Anxiety

50
Positive depression screen (%)

40

30

20

10
Time 1 Time 2 Time 3 Time 4
Training as usual
Low practice
High practice
(a)
60
Positive anxiety screen (%)

45

30

15
Time 1 Time 2 Time 3 Time 4
Training as usual
Low practice
High practice
(b)
10
Positive sleep problems screen (%)

4
Time 1 Time 2 Time 3 Time 4
Training as usual
Low practice
High practice
(c)

Figure 2: Percent of trainees’ meeting screening criteria for depression, anxiety, and sleep problems over time by interaction of embedded
practice frequency and condition.

practice at least three days per week. Given that this study moments during the day beyond a single 15-minute session.
employed a formalized group practice intervention in line It may be that when soldiers chose to embed mindfulness
with previous research [22, 23], these findings add to exist- training, they were also identifying what they could control
ing knowledge by demonstrating that mindfulness training in a context in which many demands are outside of their
can be strengthened through frequent individual embedded control. Learning how to distinguish between what one can
practice in the workplace. Study results suggest that it may and cannot control may be particularly useful in a high-
be useful to integrate mindfulness practice into smaller stress environment like the military. Importantly, future
Depression and Anxiety 9

research should examine the benefits of integrating yoga findings are also noteworthy given that the mindfulness
postures throughout the day; unfortunately, this pattern of trainers had no or limited previous background in mindful-
behavior was not assessed in the present study. ness, suggesting that this approach may be scalable for large
A number of mechanisms may explain the more rapid organizations like the military. Future research should
decrease in positive screens for depression and sleep prob- explore how the use of embedded practices can be supported
lems in the intervention group. Focusing attention on one’s in the workplace and how other outcomes related to person-
present-moment experience from a nonjudgmental perspec- nel well-being might be impacted. Given that individual
tive could help to decrease self-criticism and improve mood embedded practices were conducted in combination with
[48]. Mindfulness training may also mitigate cognitive pro- formal group practice, future research should examine the
cesses that interfere with sleep such as rumination [20, 21]. relative unique benefit of each. Mindfulness and yoga may
Additionally, learning new skills through a mindfulness enable personnel in high-stress occupations, such as health
training and postural yoga program could support mental care and first responders, to sustain their mental health even
health through increased self-esteem and a reduction in the in the context of significant psychological demands.
frequency or intensity of negative thoughts [49].
While positive screens for anxiety decreased over time
for all participants, there were no significant differences
Data Availability
between conditions. The absence of a significant time-by- The data that support the findings of this study are not cur-
condition or time-by-practice finding for anxiety in the pres- rently publicly available due institutional regulations pro-
ent study aligns with previous research on postural yoga tecting service member survey responses but are available
[13]; however, findings are not consistent with previous from the corresponding author on reasonable request (may
research showing that mindfulness training reduced anxiety require data use agreements to be developed).
symptoms in some occupational settings [19]. It may be that
the reduction in stress soldiers typically experience over the
course of BCT [50, 51] accounted for the absence of differ- Disclosure
ences by training condition. Interestingly, anxiety was
Material has been reviewed by the Walter Reed Army Institute
endorsed by the majority of study participants and the most
of Research. There is no objection to its presentation and/or
commonly reported mental health concern. For most sol-
publication. The opinions or assertions contained herein are
diers, anxiety may be fueled by the series of different and
the private views of the author and are not to be construed
unique demands encountered during BCT. Stress theory
as official, or as reflecting true views of the Department of
posits that novelty is a primary driver of anxiety [52]. Thus,
the Army or the Department of Defense. The investigators
it may be that soldiers acclimated to the environment over
have adhered to the policies for protection of human subjects
time, and this adaptation led to reduced anxiety rather than
as prescribed in AR 70–25. Findings from this manuscript
differences associated with training condition.
were presented at the Military Health System Research
Symposium in August 2023 (https://www.researchgate.net/
6.1. Limitations. A few limitations to the study merit discus-
publication/374333481_Bolstering_Soldier_Mental_Health_
sion. First, we relied on self-report for assessing mental
in_Basic_Combat_Training_with_Mindfulness_and_Yoga_A
health. Second, frequency of individual embedded yoga
_Randomized_Controlled_Trial). Findings from this
practice was not tracked, so the impact of embedded yoga
research on other outcomes have been published else-
practice outside of PRT is unknown. Third, all soldiers
where (https://www.frontiersin.org/articles/10.3389/fpsyg.20
assigned to the intervention condition received both mind-
23.1214039/full).
fulness training and postural yoga. Given that many mind-
fulness programs include postural yoga components [11,
53] and postural yoga programs often incorporate elements Conflicts of Interest
of mindfulness [13, 14], mindfulness and yoga might sepa-
rately contribute to mental health, have a synergistic effect, Amishi Jha is a coauthor and a copyright holder of the
or have differential impacts on various outcomes. Future MBAT Program. The authors have no other conflicts of
research should differentiate between the impacts of mind- interest to declare.
fulness training and postural yoga on mental health.
Acknowledgments
7. Conclusions
We want to acknowledge the assistance of Paul E. Funk II,
To our knowledge, the present study represents the largest Kevin Bigelman, Kimberly Jordan, Tony Best, Alexis Dixon,
systematic effort to explore the impact of mindfulness and Peter Mikoski, Yvonne Allard, Richard Gonzales, Stephanie
yoga on mental health among personnel in a high-risk occu- Bricault, Joy Hocut, Gery Denniswara, Molly Schwalb, Sid-
pation. Our findings provide evidence for the positive hartha Chaudhury, Kwontiera Dabney, Na Hyun Lee, Jae-
impact of a combined mindfulness and yoga intervention sarr De Guzman, and Scott Rogers for their support. This
on depression and sleep problems among soldiers during work was supported by the U.S. Army Training and Doc-
high-stress training and highlight the added benefits of trine Command (TRADOC) and the Military Operational
embedding mindfulness practice into everyday life. Positive Medicine Research Program (MOMRP).
10 Depression and Anxiety

Supplementary Materials Kabat-Zinn, Thich Nhat Hanh: 8601300234281: Amazon.

com: BooksAugust 2022, https://www.amazon.com/Full-
Figure S1 shows percentages and confidence interval bands Catastrophe-Living-Revised-Illness/dp/0345536932.
for positive screens for depression, anxiety, and sleep prob- [13] H. Cramer, R. Lauche, D. Anheyer et al., “Yoga for anxiety: a
lems over time-by-condition. Figure S2 depicts percentages systematic review and meta-analysis of randomized controlled
and confidence interval bands for positive screens for trials,” Depression and Anxiety, vol. 35, no. 9, pp. 830–843, 2018.
depression, anxiety, and sleep problems over time by inter- [14] E. J. Groessl, M. Maiya, A. R. Elwy et al., “The essential prop-
action of embedded practice frequency and condition. Con- erties of yoga questionnaire: development and methods,”
fidence interval bands in each figure were computed using International Journal of Yoga Therapy, vol. 25, no. 1, pp. 51–
Wilson’s confidence interval for binomial proportions. 59, 2015.
(Supplementary Materials) [15] A. P. Jha, E. A. Stanley, A. Kiyonaga, L. Wong, and L. Gelfand,
“Examining the protective effects of mindfulness training on
References working memory capacity and affective experience,” Emotion,
vol. 10, no. 1, pp. 54–64, 2010.
[1] Y. Li, N. Scherer, L. Felix, and H. Kuper, “Prevalence of depres- [16] A. P. Jha, A. P. Zanesco, E. Denkova et al., “Bolstering cogni-
sion, anxiety and post-traumatic stress disorder in health care tive resilience via train-the-trainer delivery of mindfulness
workers during the COVID-19 pandemic: a systematic review training in applied high-demand settings,” Mindfulness,
and meta-analysis,” PLoS One, vol. 16, no. 3, article e0246454, vol. 11, no. 3, pp. 683–697, 2020.
2021. [17] R. A. Baer, “Mindfulness training as a clinical intervention: a
[2] W. V. McCall, D. Mensah-Bonsu, A. E. Withers, and R. W. conceptual and empirical review,” Clinical Psychology: Science
Gibson, “Short-term insomnia disorder in health care workers and Practice, vol. 10, no. 2, pp. 125–143, 2003.
in an academic medical center before and during COVID-19: [18] B. Khoury, T. Lecomte, G. Fortin et al., “Mindfulness-based
rates and predictive factors,” Journal of Clinical Sleep Medi- therapy: a comprehensive meta-analysis,” Clinical Psychology
cine, vol. 17, no. 4, pp. 749–755, 2021. Review, vol. 33, no. 6, pp. 763–771, 2013.
[3] S. Jones, “Describing the mental health profile of first [19] M. Janssen, Y. Heerkens, W. Kuijer, B. Van Der Heijden, and
responders: a systematic review,” Journal of the American Psy- J. Engels, “Effects of mindfulness-based stress reduction on
chiatric Nurses Association, vol. 23, no. 3, pp. 200–214, 2017. employees’ mental health: a systematic review,” PLoS One,
[4] A. A. Vujanovic, A. Lebeaut, and S. Leonard, “Exploring the vol. 13, no. 1, article e0191332, 2018.
impact of the COVID-19 pandemic on the mental health of [20] J. A. Rash, V. A. J. Kavanagh, and S. N. Garland, “A meta-
first responders,” Cognitive Behaviour Therapy, vol. 50, no. 4, analysis of mindfulness-based therapies for insomnia and
pp. 320–335, 2021. sleep disturbance: moving towards processes of change,” Sleep
[5] B. A. Moore, L. M. Tison, J. G. Palacios, A. L. Peterson, and Medicine Clinics, vol. 14, no. 2, pp. 209–233, 2019.
V. Mysliwiec, “Incidence of insomnia and obstructive sleep [21] N. Y. Winbush, C. R. Gross, and M. J. Kreitzer, “The effects of
apnea in active duty United States military service members,” mindfulness-based stress reduction on sleep disturbance: a
Sleep, vol. 44, no. 7, article zsab024, 2021. systematic review,” Explore, vol. 3, no. 6, pp. 585–591, 2007.
[6] M. S. Dunbar, M. S. Schuler, S. O. Meadows, and C. C. Engel, [22] T. H. Nassif, A. L. Adrian, I. A. Gutierrez et al., “Optimizing
“Associations between mental and physical health conditions performance and mental skills with mindfulness-based atten-
and occupational impairments in the U.S. military,” Military tion training: two field studies with operational units,” Military
Medicine, vol. 187, no. 3–4, pp. e387–e393, 2022. Medicine, vol. 188, no. 3-4, pp. e761–e770, 2023.
[7] C. Madsen, M. Vaughan, and T. P. Koehlmoos, “Use of inte- [23] A. P. Zanesco, E. Denkova, S. L. Rogers, W. K. MacNulty, and
grative medicine in the United States Military Health System,” A. P. Jha, “Mindfulness training as cognitive training in high-
Evidence-Based Complementary and Alternative Medicine, demand cohorts: an initial study in elite military servicemem-
vol. 2017, Article ID 9529257, 11 pages, 2017. bers,” Progress in Brain Research, vol. 244, pp. 323–354, 2019.
[8] V. F. Capaldi, J. R. Kim, A. A. Grillakis, M. R. Taylor, and C. M. [24] P. Blanck, S. Perleth, T. Heidenreich et al., “Effects of mindful-
York, “Insomnia in the military: application and effectiveness ness exercises as stand-alone intervention on symptoms of
of cognitive and pharmacologic therapies,” Current Psychiatry anxiety and depression: systematic review and meta-analysis,”
Reports, vol. 17, no. 10, p. 85, 2015. Behaviour Research and Therapy, vol. 102, pp. 25–35, 2018.
[9] J. A. Dopheide, “Insomnia overview: epidemiology, patho- [25] R. B. Domingues, “Modern postural yoga as a mental health
physiology, diagnosis and monitoring, and nonpharmacologic promoting tool: a systematic review,” Complementary Thera-
therapy,” The American Journal of Managed Care, vol. 26, pies in Clinical Practice, vol. 31, pp. 248–255, 2018.
Supplement 4, pp. S76–S84, 2020. [26] L. Louie, “The effectiveness of yoga for depression: a critical lit-
[10] P. D. Bliese, A. B. Adler, and C. A. E. Castro, Eds., “Research- erature review,” Issues in Mental Health Nursing, vol. 35, no. 4,
based preventive mental health care strategies in the military,” pp. 265–276, 2014.
in Deployment psychology: Evidence-based strategies to pro- [27] N. K. Manjunath and S. Telles, “Influence of yoga and Ayur-
mote mental health in the military, pp. 103–124, American veda on self-rated sleep in a geriatric population,” Indian Jour-
Psychological Association, Washington, DC, USA, 2011. nal of Medical Research, vol. 121, no. 5, pp. 683–690, 2005.
[11] R. S. Crane, J. Brewer, C. Feldman et al., “What defines [28] L. M. Puerto Valencia, A. Weber, H. Spegel et al., “Yoga in the
mindfulness-based programs? The warp and the weft,” Psycho- workplace and health outcomes: a systematic review,” Occupa-
logical Medicine, vol. 47, no. 6, pp. 990–999, 2017. tional Medicine, vol. 69, no. 3, pp. 195–203, 2019.
[12] Full Catastrophe Living (Revised Edition): Using the Wisdom of [29] S. Telles, R. K. Gupta, S. Verma, N. Kala, and A. Balkrishna,
Your Body and Mind to Face Stress, Pain, and Illness: Jon “Changes in vigilance, self rated sleep and state anxiety in
Depression and Anxiety 11

military personnel in India following yoga,” BMC Research [46] E. A. Hoge, R. L. Acabchuk, H. Kimmel et al., “Emotion-
Notes, vol. 11, no. 1, p. 518, 2018. related constructs engaged by mindfulness-based interven-
[30] S. Telles, N. Kala, R. K. Gupta et al., “Effect of yoga on vigi- tions: a systematic review and meta-analysis,” Mindfulness,
lance, self rated sleep and state anxiety in border security force vol. 12, no. 5, pp. 1041–1062, 2021.
personnel in India,” Work, vol. 63, no. 2, pp. 243–251, 2019. [47] M. A. Khusid and M. Vythilingam, “The emerging role of
[31] E. Denkova, A. P. Zanesco, S. L. Rogers, and A. P. Jha, “Is resil- mindfulness meditation as effective self-management strategy,
ience trainable? An initial study comparing mindfulness and part 1: clinical implications for depression, post-traumatic
relaxation training in firefighters,” Psychiatry Research, stress disorder, and anxiety,” Military Medicine, vol. 181,
vol. 285, article 112794, 2020. no. 9, pp. 961–968, 2016.
[32] Pocket yoga, Rainfrog, LLC, 2022, https://pocketyoga.com/ [48] L. A. Uebelacker and M. K. Broughton, “Yoga for depression
pose/?sort=category. and anxiety: a review of published research and implications
[33] “A-Z directory of yoga poses,” Yoga Journalhttps://www for healthcare providers,” Focus, vol. 16, no. 1, pp. 95–97, 2018.
.yogajournal.com/pose-finder/. [49] T. L. Da Silva, L. N. Ravindran, and A. V. Ravindran, “Yoga in
[34] Army Regulation 70-25: use of volunteers as subjects of research, the treatment of mood and anxiety disorders: a review,” Asian
Department of the Army, 1990, https://armypubs.army.mil/ Journal Psychiatry, vol. 2, no. 1, pp. 6–16, 2009.
epubs/DR_pubs/DR_a/pdf/web/r70_25.pdf. [50] A. B. Adler, J. Williams, D. McGurk, A. Moss, and P. D. Bliese,
[35] K. Kroenke, R. L. Spitzer, and J. B. W. Williams, “The patient “Resilience training with soldiers during basic combat train-
health questionnaire-2,” Medical Care, vol. 41, no. 11, ing: randomisation by platoon,” Applied Psychology: Health
pp. 1284–1292, 2003. and Well-Being, vol. 7, no. 1, pp. 85–107, 2015.
[36] L. G. Staples, B. F. Dear, M. Gandy et al., “Psychometric prop- [51] H. R. Lieberman, J. P. Karl, J. P. McClung, K. W. Williams, and
erties and clinical utility of brief measures of depression, anx- S. Cable, “Improved mood state and absence of sex differences
iety, and general distress: the PHQ-2, GAD-2, and K-6,” in response to the stress of Army Basic Combat Training,”
General Hospital Psychiatry, vol. 56, pp. 13–18, 2019. Applied Psychology: Health and Well-Being, vol. 8, no. 3,
[37] Y.-L. Zhang, W. Liang, Z.-M. Chen et al., “Validity and reli- pp. 351–363, 2016.
ability of patient health questionnaire-9 and patient health [52] R. S. Lazarus and S. Folkman, Stress, Appraisal, and Coping,
questionnaire-2 to screen for depression among college stu- Springer Publishing Company, 1984.
dents in China,” Asia-Pacific Psychiatry, vol. 5, no. 4, [53] S. Wagh-Gumaste, “Influence of Hindu spiritual teachers on
pp. 268–275, 2013. mindfulness-based stress reduction (MBSR) of Jon Kabat-
[38] K. Kroenke, R. L. Spitzer, J. B. W. Williams, P. O. Monahan, Zinn: focusing on the teachings of Sri Nisargadatta Maharaj
and B. Löwe, “Anxiety disorders in primary care: prevalence, and Sri Ramana Maharshi,” International Journal of South
impairment, comorbidity, and detection,” Annals of Internal Asian Studies, vol. 12, pp. 1–18, 2022.
Medicine, vol. 146, no. 5, pp. 317–325, 2007.
[39] F. Plummer, L. Manea, D. Trepel, and D. McMillan, “Screen-
ing for anxiety disorders with the GAD-7 and GAD-2: a sys-
tematic review and diagnostic metaanalysis,” General
Hospital Psychiatry, vol. 39, pp. 24–31, 2016.
[40] C. H. Bastien, A. Vallières, and C. M. Morin, “Validation of the
insomnia severity index as an outcome measure for insomnia
research,” Sleep Medicine, vol. 2, no. 4, pp. 297–307, 2001.
[41] A. B. Adler, B. C. Gunia, P. D. Bliese, P. Y. Kim, and M. L.
LoPresti, “Using actigraphy feedback to improve sleep in sol-
diers: an exploratory trial,” Sleep Health, vol. 3, no. 2,
pp. 126–131, 2017.
[42] H. Chakraborty and G. Hong, “A mixed model approach for
intent-to-treat analysis in longitudinal clinical trials with miss-
ing values,” in RTI Press Methods Report Series, RTI Press,
Research Triangle Park (NC), 2009, July 2023, http://www
.ncbi.nlm.nih.gov/books/NBK538904/.
[43] J. Spybrook, S. W. Raudenbush, X.-f. Liu, R. Congdon, and
A. Martínez, “Optimal design for longitudinal and multilevel
research: documentation for the ‘Optimal Design’ software,”
2006, July 2023, https://www.semanticscholar.org/paper/
Optimal-Design-for-Longitudinal-and-Multilevel-for-
S p y b r o o k - R a u d e n b u s h /
968ee295121b114db964b384456237e7d7263e0b.
[44] R Core Team, A language and environment for statistical com,
R Foundation for Statistical Computing, Vienna, Austria,
2021, https://www.R-project.org/.
[45] D. Bates, M. Mächler, B. Bolker, and S. Walker, “Fitting linear
mixed-effects models using lme4,” 2014, https://arxiv.org/1406
.5823.
Copyright of Depression & Anxiety (1091-4269) is the property of Hindawi Limited and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.