Content uploaded by Saratu Garba Abdullahi
Author content
All content in this area was uploaded by Saratu Garba Abdullahi on Mar 15, 2023
Content may be subject to copyright.
Open camera or QR reader and
scan code to access this article
and other resources online.
Effects of Online Mindfulness-Based
Interventions on Depressive Symptoms
in College and University Students:
A Systematic Review and Meta-Analysis
Chuntana Reangsing, PhD, RN,
1
Saratu Garba Abdullahi, RN,
2
and Joanne Kraenzle Schneider, PhD, RN
2
Abstract
Introduction: Depression is considered a multiproblematic disorder that leads to impairment in interpersonal,
academic, social, and occupational functioning. Untreated depression can lead to suicide, which is the second
leading cause of death among adolescents and young adults. Antidepressants and psychotherapy have limited
effectiveness and are not available worldwide. Alternative and complementary treatments, such as online
mindfulness-based interventions (MBIs), are growing.
Objective: We examined the effects of online MBIs on depressive symptoms in college and university
students and explored the moderating effects of participant, methods, and intervention characteristics.
Methods: We systematically searched nine databases from their inception through August 2022 without date
restrictions. We included primary studies evaluating MBIs with college and university students with depression
measured as an outcome, a comparison group, that were written in English. We used random-effects model to
compute effect sizes (ESs) using Hedges’ g, a forest plot, and Qand I
2
statistics as measures of heterogeneity;
we also examined moderator analyses.
Results: Fifteen studies included 1886 participants (22.6 –3.2 years old). Overall, online MBIs showed
significantly improved depression (g=0.18, 95% confidence interval 0.02 to 0.34, I
2
=61%) compared with
controls. With regard to moderators, when depression was measured further from the end of the interven-
tion, there was less reduction in depressive symptoms (b=-0.012, Q
model
=3.81, p=0.051). Researchers who
reported higher attrition reported less beneficial effects on depressive symptoms (b=-0.013, Q
model
=9.85,
p=0.001). Researchers who used intention-to-treat reported lower ESs (g=-0.15) compared with not using
intention-to-treat (g=0.32, p<0.001). No other quality indicators moderated the effects of online MBIs on
depression.
Conclusions: Online MBIs improved depressive symptoms in college and university students. Thus, it might
be used as one treatment in their tool kit for college and university students.
Keywords: depression, college and university students, mindfulness, meta-analysis
1
School of Nursing, Mae Fah Luang University, Chiang Rai, Thailand.
2
Trudy Busch Valentine, School of Nursing, Saint Louis University, Saint Louis, MO, USA.
JOURNAL OF INTEGRATIVE AND COMPLEMENTARY MEDICINE JICM
Volume 00, Number 00, 2022, pp. 000–000
ªMary Ann Liebert, Inc.
DOI: 10.1089/jicm.2022.0606
1
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
Introduction
Approximately 80% of all mental health problems
such as depression appear before 25 years of age, which
is a critical period for adolescents and young adults.
1,2
It is a
time of multiple changes in roles, expectations, and respon-
sibilities,
3
especially for college and university students who
need to balance their educational and social lives.
4
Depres-
sion prevalence varies across and within countries. Li et al
5
reported that the prevalence of depression ranged from 7.9%
to 40.1%, with a weighted mean prevalence of 33.6% (95%
confidence interval [CI]: 29.3 to 37.8). Similarly, Ochnik
et al
6
found that about 40% of university students across nine
countries (Poland, Slovenia, Czechia, Ukraine, Russia, Ger-
many, Turkey, Israel, and Columbia) had experienced de-
pressive symptoms. Also, Capdevila-Gaudens et al
7
reported
that about 41% of Spanish medical students had depressive
symptoms and 23% of those had moderate to severe levels.
7
This is similar to the United States where Casey et al
8
found
that about 42% of the U.S. college students had moderate
symptoms of depression.
Depression is considered to be common mental health
disorder that leads to impairment in interpersonal, academic,
social, and occupational functioning.
4
Untreated depression
can lead to suicide, which is the second leading cause of
death among adolescents and young adults. Interestingly,
about 13% of college students reported suicide ideation in
the past year.
8
Currently, the number of psychological services on uni-
versities has grown; university counseling services are
dealing with increasingly severe mental health illnesses
among students. However, there are several barriers such as
cost of treatment, lack of mental health professionals, and
personal obstacles (stigma, negative perception of treat-
ment) that play important roles in the access of treatment.
Thus, alternative and complementary therapies to improve
depressive symptoms for college and university students are
growing. One of these therapies is online mindfulness-based
interventions (online MBIs).
Mindfulness is defined as paying attention on purpose to
the present moment with nonjudgment.
9
During mindfulness
practice, individuals become more aware of their thoughts,
emotions, and sensations. Notably, mindfulness training is
recognized as a cognitive training because individuals are
encouraged to understand the relationship between their
thoughts, emotions, and behaviors related to depression.
With mindfulness, they train their minds to be more flexible,
reasonable, and positive. Thus, practicing mindfulness can
decrease the severity of depressive symptoms.
Typically, online MBIs include mindfulness-based stress
reduction (MBSR), which focuses on stress management
strategies combined with mindfulness meditation, body
awareness, and yoga.
9
In contrast, mindfulness-based cog-
nitive therapy (MBCT) was designed for mentally and
physically ill patients to learn about the relationship be-
tween thoughts, emotions, and behaviors, and teach them to
become more flexible, reasonable, and positive.
10,11
For
adapted MBIs, researchers adapt the components of their
mindfulness intervention to their particular population.
12,13
Meta-analyses researchers reported that online MBIs have
beneficial effects on depression in both clinically and non-
clinically depressed patients,
14,15
but these meta-analyses
have considerable limitations. For example, Spijkerman
et al
14
reported that online MBIs had a small effect on de-
pression in the general population (g=0.29, 95% CI: 0.13 to
0.46, p=0.001, s=12). However, only 3 of the 12 primary
studies focused on students, and they did not meta-analyze
those studies separately. In another meta-analysis, Li et al
15
found that MBIs were effective in decreasing depressive
symptoms in nursing students after 8-week interventions
(standardized mean difference =-0.70, 95% CI: -1.14 to
-0.26, p=0.002).
However, the number of included primary studies was
small (S=2), and they did not specifically examine MBIs
delivered online. Importantly, no published meta-analysis
targets the specific effects of online-delivered MBI (online
MBIs) on depressive symptoms in college and university
students. Thus, the objective of this study was to examine
the effect of online MBIs on depressive symptoms in college
and university students. Also, we explored the moderator
effects of source, participant, methods, and intervention
characteristics. We hypothesized that college and university
students who engaged in online MBIs would have fewer
depressive symptoms than college and university students
who did not engage in online MBIs.
Methods
This study was conducted in accordance with the Pre-
ferred Reporting Items for Systematic reviews and Meta-
Analyses (PRISMA) statement for reporting systematic
reviews and meta-analyses of primary studies.
16
Search strategy and selection criteria
We conducted a systematic literature search across nine
electronic databases in collaboration with an expert librari-
an: Scopus (1788+), CINAHL (1937+), PubMed (1809+),
Ovid MEDLINE (1946+), Ovid PsycINFO (1967+), Co-
chrane Library (1995+), ProQuest Dissertation and Theses
(1996+), ScienceDirect (1880+), and Mindfulness-Journal
Springer (2010+). Each database was initially searched for
English language journal articles from the inception date
(in parentheses) through August 15, 2022, using the fol-
lowing search terms: (mindful* OR meditat*) AND ((mo-
bile AND (program OR intervention)) OR (smartphone
app*) OR (web* intervention) OR (online intervention) OR
(ehealth OR mHealth OR telehealth OR (internet-deliver*)
AND (program OR intervention))). Truncating terms with
an asterisk includes all forms of the terms. We exploded
subject headings (Supplementary Table S1).
Selection of studies
After the removal of duplicates, we reviewed the re-
maining titles and screened the abstracts for potentially
relevant articles. Finally, we obtained the full text of the
selected articles and assessed them for eligibility. The
screening of titles and abstracts and full-text articles was
independently conducted by two authors (C.R. and A.S.G.).
Disagreements were discussed until consensus was reached.
To conduct moderator analyses, we opted for rather broad
inclusion criteria and coded those differences across studies.
We included primary studies that: (1) employed MBIs (in-
cluding MBSR, MBCT, adapted MBIs) either with or
2 REANGSING ET AL.
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
without guided meditation; (2) administer the MBIs via
internet or a computer application (including virtual class-
rooms); (3) measured depressive symptoms as a quantitative
outcome; (4) administered the MBIs to college or university
students; and (5) used a control or comparison group
(waitlist control/control group or usual care group).
We excluded studies in which interventions were merely
a psychoeducational program and did not involve mind–
body exercise for enhancing mindfulness. We also excluded
studies in which the researchers combined MBI and other
forms of therapy (e.g., cognitive behavioral therapy, sup-
portive therapy, antidepressant treatment or therapies such
as yoga, Tai Chi, qigong, transcendental meditation, TM;
acceptance and commitment therapy, ACT; or dialectical
behavior therapy, DBT), making it difficult to disentangle
the effect of online MBI from the other therapies, because
we were specifically interested in the effect of online
MBIs on depression. We also excluded primary studies
with less than four participants in each group. Two au-
thors were contacted by e-mail and they both provided
additional information originally missing from their pri-
mary studies.
Data extraction and coding
We developed a codebook to extract data from the eli-
gible studies across five categories, that is, source, methods,
interventions, participants, and outcomes. Source variables
included the eligibility criteria and the author, year, funding
source, country, and publication status. Method variables
included setting, type of comparison group, sampling, and
quality indicators such as group assignment, conceal allo-
cation, masked data collectors, intention-to-treat analysis,
fidelity check, power estimation, and group comparisons.
Intervention variables included the type (MBSR, MBCT,
MBI); format (mobile application, website-based interven-
tion); whether or not the intervention was guided or included
music, body scan, psychoeducation, group discussion, sit-
ting meditation, body movement, and whether there was
counseling or home assignments.
We also extracted days across the intervention, number
of weeks in which the intervention was administered,
number of intervention sessions, and minutes per session.
Participant variables included the total number of the par-
ticipants enrolled, their mean age and standard deviation,
number of participants in the intervention group and in the
control group, number of participants at analysis in both
groups, number of dropouts, number of females, number of
participants across races, and presence of mood disorder,
anxiety disorder, stress, learning disorder, and the use of
drugs. Finally, the outcome variables included the name of
the depression instruments, scale reliabilities, means and
standardized deviations of scale scores, and the effect
direction.
Data were extracted by two independent researchers (C.R.
and A.S.G.). Data were entered and managed utilizing the
Research Electronic Data Capture (REDCap) database appli-
cation.
17
REDCap is a system for structured, clinical study
data capture, and is designed to comply with HIPAA regula-
tions. Any inconsistencies in data extraction were resolved via
discussion between two researchers (C.R. and A.S.G.) and
through consultation with the third researcher ( J.K.S.).
Statistical analysis
We used SPSS to conduct descriptive statistics of the
study characteristics. We used Comprehensive Meta-
Analysis (CMA) version 3.0 to compute the effect size (ES)
by computing the standardized mean differences between
online MBIs and comparison groups’ post-test depression
scores. Using standardized mean differences allowed us to
compare effects that were measured with instruments of
different metrics. Because studies differ in unmeasurable
ways such as intervention delivery, participant mixes, set-
ting features, and more, we assumed that the studies had
different underlying true ESs. Thus, we used random-effects
model because we assumed that the true effects were nor-
mally distributed. Using the random-effects model, CMA
weights each study by the inverse of the within- and
between-studies variance to estimate the mean of a distri-
bution of true effects, so the weights are more moderate
compared with fixed-effect models.
18
To estimate the ES,
we used Hedges’ gcorrection for small study samples with
95% confidence intervals (CIs).
19,20
Heterogeneity assessment
We tested the heterogeneity across studies using the forest
plot, which visually demonstrates the degree to which data
from multiple studies overlap with one another. Also, we
used the Qstatistic to indicate the total dispersion; signifi-
cance indicates heterogeneity. Additionally, we used the I
2
statistic, which is the ratio of ES variability to total vari-
ability indicating the proportion of variance in true effects,
that is, the proportion of observed variance that reflects real
differences.
21
It reflects how much the observed study ESs
differ from each other than what we would have expected
due to chance alone.
22
The I
2
reflects the proportion of
variance that is true.
20
A value of 25%, 50%, and 75%
reflect low, moderate, and high variability, respectively.
21
When heterogeneity existed, we examined subgroup
analysis, that is, moderator analyses,
21
based on source,
participants, method, and intervention characteristics to
explore how these various characteristics influenced ES. We
used a meta-analytic analog of analysis of variance for
categorical moderators and meta-regression, an analog of
regression analysis for continuous moderators.
21
When the
number of primary studies was less than six in subgroups
that were significantly different, we used the Hartung,
Knapp, Sidik, and Jonkman (HKSJ) adjustment to compute
a CI for average effect.
23,24
The HKSJ adjustment modifies
the standard error of the mean and multiplies it by the t
distribution instead of the Zdistribution used in the standard
analysis. The HKSJ adjustment yields a wider and more
accurate CI when there are a small number of studies in the
analysis
22
and therefore a more accurate inference for the
average effect than the conventional method.
23
Assessment of the methodological quality
To assess the methodological quality of primary studies,
we used the quality indicators as moderators and examined
the difference in ESs for studies with and without the
quality indicators. For this meta-analysis, quality indicators
included concealed allocation, random assignment, data
collector blinded, a priori power analysis, power analysis
ONLINE MBISAND DEPRESSION IN COLLEGE AND UNIVERSITY STUDENTS 3
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
completed, group demographics compared, and intention-
to-treat analysis. These indicators were analyzed as di-
chotomous moderators, whereas attrition was analyzed as a
continuous moderator.
Estimation of bias
Additionally, we explored design bias by examining
single-group pre- and post-test comparisons to explore the
possibility of spontaneous recovery of depressive symp-
toms.
19,21,22
A significant pre–post finding in the control
group indicates that spontaneous recovery likely exists and
can be helpful in interpreting the intervention findings.
25
Although pre- and post-intervention depression scores are
likely correlated, few primary researchers report these cor-
relations. Thus, we computed pre–post ESs in two ways,
using a strong correlation (r=0.8) to be conservation and
using no correlation (r=0.0).
25
To estimate the publication bias, we used the funnel plot,
Begg and Mazumdar rank correlation test, and Egger’s
bias value. A visually symmetrical funnel-shaped distribu-
tion represents the absence of publication bias; conversely,
asymmetry suggests publication bias. The Begg and Ma-
zumdar test computes the rank order correlation (Kendall’s
tau) between the standard treatment effect and the variance
(standard error, which is primarily affected by the sample
size). Significant results for either test suggest publication
bias.
26
Ethical approval
This meta-analysis does not require ethical approval or
patient consent because the data used were extracted from
primary studies; no human subjects’ data were used.
Results
Demographics of the study
Initial database searches and updated electronic database
searches resulted in 5594 articles. After 1681 duplicates
were removed, 3913 studies remained. We found two arti-
cles through hand ancestry searches. During review of ab-
stracts, an additional 3865 were excluded because they did
not include online MBIs and/or any number of the inclusion
criteria. Of the remaining 50 studies, 37 studies were ex-
cluded: 14 were narrative review/systematic review/meta-
analyses, 17 were qualitative studies, and 6 were research
protocols. Finally, 15 primary studies (S=15) met the in-
clusion criteria and were included in this systematic review
and a meta-analysis (Fig. 1).
27–41
The 15 primary studies (S=15) that met the inclusion
criteria provided 18 between-group comparisons (K=18)
because some studies had 3 comparison groups. For exam-
ple, when primary researchers included three groups such as
MBI headspace, smiling mind, and a control,
32
we com-
pared mindfulness groups with like control group without
mindfulness intervention to examine the effects of only
online mindfulness interventions, the main objective of this
meta-analysis. All the 15 studies had been published be-
tween 2013 and 2022. A total of 1886 participants were
included across the 15 primary studies; 1028 participants
practiced in the online MBI and 858 participants served as
controls. Seven of the 15 studies were conducted in the
United States, 3 each in Canada and United of Kingdom, 2
in New Zealand, and 1 in France (Supplementary Tables S2
and S3). Participants’ mean age ranged from 18.2 to 31.1
years across studies; they were mainly female (mean 86%)
and White (mean 67%, Table 1).
Six measures were used to determine depressive symp-
toms in college and university students including the De-
pression, Anxiety, and Stress Scale-21 Items (DASS-21,
s=6), the Patient Health Questionnaire (PHQ-4/9, s=4), the
Center for Epidemiological Studies Depression (CES-D,
s=1), the Center Assessment of Psychological Symptoms-34
(CAPS-34, s=1), the Quick Inventory of Depressive
Symptomatology–Self-report (QIDS-SR, s=1), and the
Patient-Reported Outcome Measurement Information Sys-
tem (PROMIS, s=1). Higher scores reflect higher levels of
depressive symptoms. The reliability of these measures
ranged between 0.71 and 0.93 (s=10). All online MBIs
were MBIs adapted for the online environment. See Table 1
for intervention descriptions including the total weeks of
interventions, number of sessions/week, and duration of
session in min/session.
Effects of online MBIs
Figure 2 displays the forest plot of primary studies. The
size of each square represents the weight given to the study
and the influence on the summary ES. The width of each
line shows the CI of the ES in each study. In addition, the
diamond below the included studies shows the summary ES
and the width of diamond indicates the CIs for the overall
effect estimate.
19,21
The summary ES across the 18 com-
parisons was g=0.18 (95% CI: 0.024 to 0.344, p=0.024,
I
2
=61%), indicating that online MBIs had a small effect on
depressive symptoms among college and university stu-
dents. That is, college and university students who engaged
in online MBIs had lower depression scores compared with
students who did not engage in online MBIs.
Borenstein and colleagues
21
emphasized the importance
of assessing the dispersion of ESs from study to study. They
point out that when the ESs are consistent across studies, the
effect is robust and we can confidently rely on the summary
effect.
21,22
Of all 18 comparisons, only 3 comparisons had
negative effects. Importantly, the remaining 15 studies had
positive effects supporting confidence in the summary
effect.
Online MBI group pre–post comparisons demonstrated
significant reductions in depression with ESs of g=0.28
(p<0.001) for correlated groups (r=0.8) and g=0.26
(p<0.001) for uncorrelated groups. The control group pre–
post ESs showed no improvement on depressive symptoms
for both the correlated, g=0.04 ( p=0.420), and uncorrelated
groups, g=0.05 ( p=0.438), suggesting that spontaneous re-
covery of depressive symptoms was unlikely (Table 2).
Subgroup analyses
Significant heterogeneity existed across the studies (I
2
=
61%, Q=44.48, p<0.001), indicating that moderator anal-
ysis was warranted. Three variables were significant mod-
erators (Tables 3 and 4). As expected, when primary
researchers used intention-to-treat analysis, they showed lower
ESs (g=-0.15) than researchers who did not use intention-to-
treat analysis (g=0.32, p<0.001) see Supplementary Figure S1.
4 REANGSING ET AL.
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
FIG. 1. The Preferred Re-
porting Items for Systematic
reviews and Meta-Analyses
flow.
Table 1. Characteristics of the Primary Studies
Characteristic sMinimum Q1 Median Q3 Maximum Mean SD
Mean age (years) 13 18.2 20.3 21.8 24.6 31.1 22.6 3.2
Total sample size at analysis 15 16.0 58.0 96.0 162.8 205.0 104.9 55.8
MBI group 15 6.0 33.0 52.0 77.0 122.0 57.1 33.2
Control group 15 10.0 30.3 38.5 65.8 103.0 47.7 26.6
% Female 7 63.2 81.4 87.9 95.8 100.0 86.2 12.0
% White 3 34.1 34.1 68.1 — 100.0 67.4 33.0
% African American 3 0 0 0 0 20.5 6.8 11.8
% Asian 3 0 0 6.8 — 21.2 9.4 11.0
Weeks of structured MBI 15 2.0 2.0 4.0 6.0 8.0 4.6 2.3
Days across intervention (length) 15 7.0 11.5 21.5 49.3 61.0 28.3 18.1
Structured MBI session/week 12 4.5 1.0 7.0 7.0 7.0 5.9 2.0
Structured MBI min/session 11 5.0 10.0 10.0 20.0 90.0 18.5 22.2
Dose (length ·duration) 11 49.0 70.0 255.0 960.0 4410.0 674.2 1180.9
Days after intervention measured 15 0 0 0 0 30.5 4.8 11.1
% Attrition, MBI group 15 2.5 11.4 22. 7 37.7 57.4 25.8 17.1
% Attrition, control group 15 0.0 7.3 18.7 28.7 41.8 18.8 12.9
MBI, mindfulness-based intervention; Q1, first quartile; Q3, third quartile; s, number of studies providing data; SD, standard deviation.
5
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
Due to the small number of primary studies in the subgroup
analysis (s=3), we used the HKSJ adjustment; findings re-
mained significant (F=12.88, p=0.003). Moreover, researchers
who measured depression further from the end of the inter-
vention showed less beneficial effects on depressive symptoms
(b=-0.012, Q
model
=3.81, p=0.051) suggesting a waning of
effects. In addition, researchers who reported higher attrition
reported less beneficial effects on depressive symptoms
(b=-0.013, Q
model
=9.85, p=0.001).
Publication bias
The funnel plot appeared asymmetrical (Fig. 3). Egger’s
test of the intercept was 1.52 and nonsignificant (95% CI:
-1.64 to 4.69, t=1.01, df =16, p=0.161); Begg and Ma-
zumdar rank correlation test indicated a nonsignificant
Kendall’s tau of 0.18 ( p=0.136), suggesting that publica-
tion bias was unlikely. However, the power of these tests are
low due to a small number of comparisons (K=18).
19,21,42
Thus, the findings should be interpreted with caution.
Discussion
Effects of online MBIs
This is the first systematic review and meta-analysis to
synthesize primary studies on the effects of online MBIs for
depressive symptoms in college and university students. We
found that online MBIs had a small, but significant, bene-
ficial effect on depressive symptoms among students com-
pared with comparison groups. We also found that this
effect was not because of spontaneous recovery. One
FIG. 2. Forrest plot of the effect of online MBIs versus control on depression in college and university students.
Table 2. Effect Size of the Online Mindfulness-Based Intervention Versus Control Groups
Comparison
MBI group
KES p(ES) 95% CI SE I
2
Qp(Q)
Online MBI vs. control groups 18 0.184 0.024 0.024 to 0.344 0.082 61.8 44.48 <0.001
Single group
Online MBI
Pre vs. post (r=0.0) 16 0.262 <0.001 0.154 to 0.370 0.055 8.6 16.4 0.356
Pre vs. post (r=0.8) 16 0.275 <0.001 0.173 to 0.377 0.052 78.6 70.1 <0.001
Control group
Pre vs. post (r=0.0) 16 0.046 0.438 -0.070 to 0.161 0.059 15.3 17.7 0.278
Pre vs. post (r=0.8) 16 0.043 0.420 -0.061 to 0.147 0.053 78.9 71.0 <0.001
CI, confidence interval; ES, effect size; K, number of comparisons; MBI, mindfulness-based intervention; Q, heterogeneity statistics; SE,
standard error.
6 REANGSING ET AL.
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
Table 3. Categorical Moderator Results for Depression Comparing
the Mindfulness-Based Intervention Versus Control Groups
Moderator KES SE Variance 95% CI Zp(Z)Q
bet
p(Q
bet
)
Source characteristics
Funding 3.59 0.058
Unfunded 10 0.065 0.10 0.01 -0.13 to 0.26 0.66 0.506
Funded 7 0.355 0.12 0.02 0.12 to 0.58 3.00 0.003
Country 6.09 0.193
United States 8 0.080 0.12 0.01 -0.15 to 0.31 0.68 0.499
United Kingdom 3 0.031 0.18 0.03 -0.32 to 0.38 0.17 0.863
New Zealand 4 0.146 0.22 0.05 -0.29 to 0.59 0.65 0.516
Canada 2 0.525 0.16 0.03 0.21 to 0.84 3.27 0.001
France 1 0.141 0.33 0.11 -0.50 to 0.78 0.43 0.666
Method characteristics
Blinded data collection 0.396 0.529
No 2 0.023 0.27 0.07 -0.50 to 0.55 0.08 0.933
Yes 16 0.202 0.08 0.01 0.03 to 0.37 2.30 0.021
Intention-to-treat 12.37 <0.001
No 13 0.321 0.08 0.01 0.17 to 0.46 4.29 <0.001
Yes 5 -0.149 0.11 0.01 -0.37 to 0.07 -1.34 0.179
Power of sample 0.00 0.963
No 5 0.261 0.18 0.03 -0.10 to 0.62 1.43 0.152
Yes 5 0.273 0.16 0.03 -0.05 to 0.59 1.68 0.094
Baseline characteristics equal across groups 2.45 0.118
No 2 0.467 0.21 0.05 0.05 to 0.88 2.19 0.028
Yes 14 0.106 0.09 0.01 -0.07 to 0.28 1.21 0.227
Fidelity 3.63 0.057
No 13 0.102 0.08 0.01 -0.06 to 0.27 1.21 0.228
Yes 5 0.429 0.15 0.02 0.14 to 0.72 2.87 0.004
Intervention characteristics
MBI format 1.51 0.681
Mobile 7 0.209 0.14 0.02 -0.06 to 0.48 1.52 0.129
Website 7 0.073 0.13 0.02 -0.18 to 0.33 0.56 0.575
Lived video conference 3 0.339 0.18 0.04 -0.03 to 0.70 1.82 0.069
Workshop 1 0.270 0.33 0.11 -0.38 to 0.92 0.82 0.415
Body movement 0.60 0.438
No 7 0.293 0.13 0.02 0.04 to 0.56 2.23 0.026
Yes 9 0.154 0.12 0.01 -0.08 to 0.39 1.28 0.200
Outcome measure
Days after intervention measured 3.589 0.058
Immediate post-MBI 15 0.256 0.09 0.01 0.09 to 0.43 2.94 0.003
Delayed follow-up 3 -0.118 0.18 0.03 -0.47 to 0.23 -0.67 0.505
CI, confidence interval; ES, effect size; K, number of comparisons; MBI, mindfulness-based intervention; Q, heterogeneity statistics; SE,
standard error.
Table 4. Continuous Moderators of the Effects of Mindfulness-Based Intervention on Depression
Moderator KSlope SE Tau
2
Q
model
p
Study characteristics
Publication year 18 0.032 0.043 0.06 0.55 0.467
Sample characteristics
Age (mean) 14 -0.002 0.025 0.04 0.00 0.947
% Women 7 0.002 0.010 0.04 0.03 0.867
Method characteristics
% Attrition 18 -0.013 0.004 0.02 9.85 0.001
Reliability of depressive instruments 15 0.123 1.374 0.04 0.01 0.928
Intervention characteristics
Intervention length (total week) 17 0.056 0.033 0.05 2.93 0.087
Online MBI sessions per week 14 -0.026 0.028 0.01 0.82 0.365
Duration of online MBI, min/session 13 0.000 0.000 0.02 1.0 0.992
Dose (length ·duration) 13 0.000 0.000 0.024 1.0 0.840
No. of days of depression was follow-up
after intervention completed
18 -0.012 0.006 0.047 3.81 0.051
K, number of comparisons; MBI, mindfulness-based intervention; Q, heterogeneity statistics; SE, standard error.
7
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
possible explanation for this significant finding is that
practicing mindfulness via online platforms helps college
and university students practice whenever they have avail-
able time.
9,14
Online MBIs might be especially helpful for
college and university students who would otherwise not
seek help out of fear they would be stigmatized. This effect
was possibly because none of the students across the pri-
mary studies were clinical samples. Depression scores of
nonclinical samples have less room for improvement than
those who are diagnosed with depression. Thus, future re-
searchers should explore the effects of online MBIs in stu-
dents with diagnosed depression.
Another possible reason for the small ESs may be due to
adherence. Adherence is especially relevant in mindfulness
training and should be a regular practice for developing
mindfulness skills, especially college and university stu-
dents who likely have no mindfulness experience.
14,43,44
Similarly, Canby et al
45
found that adherence was associated
with the effectiveness of MBI on depression. However,
some researchers found that adherence was not associated
with psychological outcomes.
46
Thus, future researchers
might explore the influence of adherence to online MBIs on
depressive symptoms.
Moderator effects
Researchers who used intention-to-treat analysis showed
lower effects than those who did not use intention-to-treat.
Intention-to-treat analysis is a conservative method of esti-
mating intervention effect because all participants are in-
cluded in the analysis based on what groups they were
originally assigned.
47
Thus, intention-to-treat analysis in-
cludes even those participants who dropped out and there-
fore had no improvement. An intention-to-treat analysis
provides an accurate (unbiased) conclusion of the effec-
tiveness of the intervention at the level of adherence in the
study.
48
Evaluating depression immediately postintervention had a
greater beneficial effect on depressive symptoms than when
evaluated at a delayed follow-up postintervention. In other
words, when depression was measured further from the in-
tervention, online MBIs had less of an effect on depressive
symptoms. One possible explanation might be that overtime
adherence to practice waned. Typically, after the interven-
tion, most primary researchers encouraged participants to
continue their practice at home. It is likely that students had
busy lives and their practice waned after the formal online
MBI sessions. Similarly, Galante et al
49
found that home
mindfulness practice had positive effects on well-being for
university students; however, few participants meditated
after completing the formal course with dedicated time.
49
As attrition increased, the effects of online MBIs reduced.
That is, for every one percent increase in attrition rate, the
effect of online MBIs reduced indicating an increase in
depressive scores. Higher attrition also results in a smaller
number of participants in the analysis than were enrolled, a
reduction in precision of ES.
50
Thus, we would recommend
that future researchers account for attrition during recruit-
ment of participants.
51
Strengths and limitations
Our study was the first comprehensive systematic review
and meta-analysis of online MBIs on depressive symptoms
specifically in college and university students. We also ex-
amined the moderator analyses to explore how each variable
FIG. 3. Funnel plot of standard error by Hedges’ g.
8 REANGSING ET AL.
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
subgroup influenced ES. However, this meta-analysis has
limitations. Notably, we included only primary studies
written in English, and therefore, relevant studies written
in other languages could have been missed (publication
bias). Also, we found that three primary studies showed
negative effects of online MBI and one of those was sig-
nificant.
39
Although online MBI for college and university
students generally had an effect of depressive symptoms,
using it as a blanket adjunctive treatment might not be
warranted. Finally, participants in the primary studies
were mainly female (86%); thus, using online MBIs for
male college and university students might show different
effects.
Implications and recommendations
This meta-analysis provides evidence for using online
MBI for college and university students to manage depres-
sive symptoms, especially in females. Nurses and health
providers might consider using online MBIs as one treat-
ment in their tool kit to reduce depression in students. Al-
though online MBIs have a small effect on depressive
symptoms, they provide an optional treatment for university
counseling centers or primary health care centers for stu-
dents with depressive symptoms or those who are at risk,
especially when there are insufficient mental health pro-
fessionals. And, online services such as MBIs might be
useful for students who are concerned about the negative
perception of depression treatment. Moreover, educators can
help students by integrating online MBIs into the educa-
tional curricula targeting the management of depressive
symptoms.
With regard to research implications, we encourage re-
searchers to report study quality indicators. Interestingly,
most research teams implemented online MBIs over short
periods of time (for eight sessions or less) and only mea-
sured outcomes immediately postintervention. Thus, future
researchers might explore the long-term effects of online
MBIs on depression. Also, researchers might consider how
to support college and university students in maintaining a
regular mindfulness practice on their own, to maintain its
benefits overtime. Moreover, future researchers might test
the effects of online MBI on depression in male college and
university students.
Conclusions
In conclusion, we found that college and university stu-
dents who engaged in online MBIs showed less depressive
symptoms than those who did not. Nurses and health pro-
viders may use an online MBI as one treatment in their
tool kit for managing depressive symptoms in college and
university students. Also, health providers might engage
high-risk students in online MBI in an effort to prevent
depression.
Compliance with Ethical Standards
Ethical approval not required because data used in this
meta-analysis comprised information available in the public
domain. Informed consent not required because data used in
this meta-analysis comprised information available in the
public domain and did not involve humans. Research in-
volving human participants and/or animals: this article does
not contain human participants or their data.
Data Availability Statement
The authors confirm that the data supporting the findings
of this study are available within the article and its Sup-
plementary Data.
Authors’ Contributions
All three authors (C.R., A.S.G., and J.K.S.) were re-
sponsible for acquisition, interpretation, and drafting the
article. All authors critically revised the work for important
intellectual content. The first author (C.R.) and the second
author (A.S.G.) were included in the identification, selec-
tion, data extraction, article drafting, and critically revised
the work. All authors (C.R., A.S.G., and J.K.S.) provided
final approval of the version to be published and agree to be
accountable for all aspect of the work.
Author Disclosure Statement
All the authors declared no potential conflicts of interest
with respect to the research, authorship, and/or publication
of this article.
Funding Information
The preparation and publication of this study was un-
funded.
Supplementary Material
Supplementary Fig S1
Supplementary Table S1
Supplementary Table S2
Supplementary Table S3
PRISMA 2009 Checklist_CR
References
1. Ward C, McLafferty M, McLaughlin J, et al. Suicidal
behaviours and mental health disorders among students
commencing college. Psychiatry Res 2022;307:114314.
2. Auerbach RP, Alonso J, Axinn WG, et al. Mental disorders
among college students in the World Health Organization
World Mental Health Surveys. Psychol Med 2016;46:
2955–2970; doi: 10.1017/S0033291716001665
3. Arnett JJ. Emerging adulthood: A theory of development
from the late teens through the twenties. Am Psychol 2000;
55(5):469–480.
4. Videbeck SL. Psychiatric and Mental Health Nursing. 8th
ed. Walter & Kruwer; Philadelphia, PA; 2020.
5. Li W, Zhao Z, Chen D, et al. Prevalence and associated factors
of depression and anxiety symptoms among college students:
A systematic review and meta-analysis. J Child Psychol
Psychiatry 2022;63(11):1222–1230; doi: 10.1111/jcpp.13606
6. Ochnik D, Rogowska AM, Kus
´nierz C, et al. Mental health
prevalence and predictors among university students in
nine countries during the COVID-19 pandemic: A cross-
national study. Sci Rep 2021;11:18644; doi: 10.1038/s41598-
021-97697-3
7. Capdevila-Gaudens P, Garcıa-Abajo JM, Flores-Funes D,
et al. Depression, anxiety, burnout and empathy among
ONLINE MBISAND DEPRESSION IN COLLEGE AND UNIVERSITY STUDENTS 9
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
Spanish medical students. PLoS One 2021;16(12):e0260359;
doi: 10.1371/journal.pone.0260359
8. Casey SM, Varela A, Marriott JP, et al. The influence of
diagnosed mental health conditions and symptoms of de-
pression and/or anxiety on suicide ideation, plan, and at-
tempt among college students: Findings from the Healthy
Minds Study, 2018–2019. J Affect Disord 2022;298:464–
471; doi: 10.1016/j.jad.2021.11.006
9. Kabat-Zinn J. Mindfulness-based interventions in context:
Past, present, and future. Clin Psychol Sci Pract 2003;
10(2):144–155; doi: 10.1093/clipsy/bpg016
10. Segal ZV, Dimidjian S, Beck A, et al. Outcomes of online
mindfulness-based cognitive therapy for patients with re-
sidual depressive symptoms: A randomized clinical trial.
JAMA Psychiatry 2020;77(6):563–573; doi: 10.1001/jama
psychiatry.2019.4693
11. Segal Z, Williams JM, Teasdale JD. Mindfulness-Based
Cognitive Therapy for Depression. The Guilford Press:
New York; 2002.
12. Kra
¨geloh CU, Medvedev ON, Taylor T, et al. A pilot
randomized controlled trial for a videoconference-delivered
mindfulness-based group intervention in a nonclinical set-
ting. Mindfulness 2019;10(4):700–711; doi: 10.1007/s12671-
018-1024-y
13. Rodriguez MA, Eisenlohr-Moul TA, Su K, et al. Using
non-specialist providers to improve treatment engagement
in an online mindfulness intervention in China. Behav Res
Ther 2020;130:103644; doi: 10.2196/25772
14. Spijkerman MP, Pots WT, Bohlmeijer ET. Effectiveness of
online mindfulness-based interventions in improving men-
tal health: A review and meta-analysis of randomised
controlled trials. Clin Psychol Rev 2016;45:102–114; doi:
10.1016/j.cpr.2016.03.009
15. Li YF, Sun WX, Sun XJ, et al. Effects of mindfulness
meditation on anxiety, depression, stress, and mindfulness
in nursing students: A meta-analysis and trial sequential
analysis of randomized controlled trials. Front Nurs 2020;
7(1):59–69; doi: 10.2478/fon-2020-0001
16. Moher D, Liberati A, Tetzlaff J, et al. The PRISMA Group.
Preferred reporting items for systematic reviews and meta-
analyses: The PRISMA statement. PLoS Med 2009;6(7):
e1000097; doi: 10.1136/bmj.b2535
17. Harris PA, Taylor R, Thielke R, et al. Research Electronic
Data Capture (REDCap)—A metadata-driven methodology
and workflow process for providing translational research
informatics support. J Biomed Inform 2009;42:377–381;
doi: 10.1016/j.jbi.2008.08.010
18. Borenstein M, Hedges LV, Higgins JPT, et al. Regression
in Meta-Analysis. 2017. Available from: https://www.meta-
analysis.com/downloads/MRManual.pdf [Last accessed:
September 21, 2022].
19. Borenstein M, Hedges LV, Higgins JPT, et al. Introduction
to Meta-Analysis. John Wiley & Sons, Ltd.: West Sussex,
UK; 2009.
20. Hedges LV, Pigott TD. The power of statistical tests for
moderators in meta-analysis. Psychol Methods 2004;9(4):
426–445.
21. Borenstein M, Hedges LV, Higgins JPT, et al. Introduction
to Meta-Analysis. 2nd ed. Wiley & Sons Ltd.: Oxford;
2021.
22. Cuijpers P, Weitz E, Cristea IA, et al. Pre-post effect sizes
should be avoided in meta-analyses. Epidemiol Psychiatr
Sci 2017;26:364–368; doi: 10.1017/S2045796016000809
23. van Aert R, Jackson D. A new justification of the Hartung-
Knapp method for random-effects meta-analysis based
on weighted least squares regression. Res Synth Methods
2019;10(4):515–527; doi: 10.1002/jrsm.1356
24. Jackson D, Law M, Ru
¨cker G, et al. The Hartung-Knapp
modification for random-effects meta-analysis: A useful
refinement but are there any residual concerns? Stat Med
2017;36:3923–3934; doi: 10.1002/sim.741
25. Conn VS, Hafdahl AR, Moore SM, et al. Meta-analysis of
interventions to increase physical activity among cardiac
subjects. Int J Cardiol 2009;133(3):307–320; doi: 10.1016/
j.ijcard.2008.03.052
26. Cooper H, Hedges LV, Valentine JC. The Handbook of
Research Synthesis and Meta-Analysis. Russell Sage Foun-
dation: New York, NY; 2009.
27. Ahmad F, El Morr C, Ritvo P, et al. An eight-week, web-
based mindfulness virtual community intervention for stu-
dents’ mental health: Randomized controlled trial. JMIR
Ment Health 2020;7(2):e15520; doi: 10.2196/15520
28. Bosso KB. The effects of mindfulness training on BDNF
levels, depression, anxiety, and stress levels of college
students. Dissertation Abstracts Int Section B Sci Eng 2020;
81(12-B):No-Specified.
29. Cavanagh K, Strauss C, Cicconi F, et al. A randomised
controlled trial of a brief online mindfulness-based inter-
vention. Behav Res Ther 2013;51(9):573–578; doi: 10.1016/
j.brat.2013.06.003
30. Cavanagh K, Churchard A, O’Hanlon P, et al. A rando-
mised controlled trial of a brief online mindfulness-based
intervention in a non-clinical population: Replication and
extension. Mindfulness 2018;9(4):1191–1205; doi: 10.1007/
s12671-017-0856-1
31. Fish MT, Saul AD. The gamification of meditation: A
randomized-controlled study of a prescribed mobile mind-
fulness meditation application in reducing college students’
depression. Simul Gaming 2019;50(4):419–435; doi: 10.1177/
104687811985
32. Flett JAM, Hayne H, Riordan BC, et al. Mobile mindful-
ness meditation: A randomised controlled trial of the effect
of two popular apps on mental health. Mindfulness 2018;
10:863–876; doi: 10.1007/s12671-018-1050-9
33. Kingston J, Becker L, Woeginger J, et al. A randomised
trial comparing a brief online delivery of mindfulness-plus-
values versus values only for symptoms of depression:
Does baseline severity matter? J Affect Disord 2020;276:
936–944; doi: 10.1016/j.jad.2020.07.087
34. Lee RA, Jung ME. Evaluation of an mHealth App (DeS-
tressify) on university students’ mental health: Pilot trial.
JMIR Ment Health 2018;5(1):e2; doi: 10.2196/mental.8324
35. Levin ME, Hicks ET, Krafft J. Pilot evaluation of the stop,
breathe & think mindfulness app for student clients on a
college counseling center waitlist. J Am Coll Health 2020;
70(1):165–173; doi: 10.1080/07448481.2020.1728281
36. El Morr CE, Ritvo P, Ahmad F, et al. Effectiveness of an 8-
week web-based mindfulness virtual community interven-
tion for university students on symptoms of stress, anxiety,
and depression: Randomized controlled trial. JMIR Ment
Health 2020;7(7):e18595; doi: 10.2196/18595
37. Sagon AL, Danitz SB, Suvak MK, et al. The mindful way
through the semester: Evaluating the feasibility of deliv-
ering an acceptance-based behavioral program online.
J Contextual Behav Sci 2018;9:36–44; doi: 10.1016/j.jcbs
.2018.06.004
10 REANGSING ET AL.
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
38. Nolan L. Impact of a Brief Mindfulness Exercise on
Practical Nursing Students’ Test Anxiety. University of
Arizona: Tucson, AZ; 2019.
39. Nguyen-Feng VN, Greer CS, Erazier P. Using online inter-
ventions to deliver college student mental health resources:
Evidence from randomized clinical trials. Psycholog Serv
2017;14(14):481–489; doi: 10.1037/ser0000154
40. Devlillers-Reolon L, Mascret N, Sleimen-Malkoun R. On-
line mindfulness intervention, mental health and attentional
abilities: A randomized controlled trial in university stu-
dents during COVID-19 lockdown. Front Psychol 2022;13:
1–12; doi: 10.3389/fpsyg.2022.889807
41. Simonsson O, Bazin O, Fisher SD, et al. Effects of an eight-
week, online mindfulness program on anxiety and depres-
sion in university students during COVID-19: A randomized
controlled trial. Psychiatry Res 2021;305:114222; doi:
10.1016/j.psychres.2021.114222
42. Idris NRN. A comparison of methods to detect publication
bias for meta-analysis of continuous data. J Appl Sci 2012;
12(13):1413–1417.
43. Torous J, Lipschitz J, Ng M, et al. Dropout rates in clini-
cal trials of smartphone apps for depressive symptoms: A
systematic review and meta-analysis. J Affect Disord 2020;
263:413–419; doi: 10.1016/j.jad.2019.11.167
44. Forbes L, Gutierrez D, Johnson SK. Investigating ad-
herence to an online introductory mindfulness program.
Mindfulness 2018;9:271–282; doi: 10.1007/s12671-017-
0772-4
45. Canby NK, Eichel K, Peters SI, et al. Predictors of out-of-
class mindfulness practice adherence during and after a
mindfulness-based intervention. Psychosom Med 2021;
83(6):655–664; doi: 10.1097/PSY.0000000000000873
46. Ribeiro L, Atchley RM, Oken BS. Adherence to practice
of mindfulness in novice meditators: Practices chosen,
amount of time practiced, and long-term effects following
a mindfulness-based intervention. Mindfulness 2018;9(2):
401–411; doi: 10.1007/s12671-017-0781-3
47. Dossing A, Tarp S, Furst DE, et al. Modified intention-to-
treat analysis did not bias trial results. J Clin Epidemiol
2016;72:66–74; doi: 10.1016/j.jclinepi.2015.11.003
48. McCoy EC. Understanding the intention-to-treat principle
in randomized controlled trials. West J Emerg Med 2017;
18(6):1075–1078; doi: 10.5811/westjem.2017.8.35985
49. Galante J, Stochl J, Dufour G, et al. Effectiveness of pro-
viding university students with a mindfulness-based in-
tervention to increase resilience to stress: 1-year follow-up
of a pragmatic randomised controlled trial. J Epidemiol
Community Health 2021;75:151–160; doi: 10.1136/jech-
2020-214390
50. Meyerowitz-Katz G, Ravil S, Arnolda L, et al. Rates of
attrition and dropout in app-based interventions for chronic
disease: Systematic review and meta-analysis. J Med In-
ternet Res 2020;22(9):e20283; doi: 10.2196/20283
51. Osho O, Owoeye O, Armijo-Olivo S. Adherence and at-
trition in fall prevention exercise programs for community
dwelling older adults: A systematic review and meta-
analysis. J Aging Phys Act 2018;26:304–326; doi: 10.1123/
japa.2016-0326
Address correspondence to:
Chuntana Reangsing, PhD, RN
School of Nursing
Mae Fah Luang University
333 M. 1, Thasud, Muang District
Chiangrai Province 57100
Thailand
E-mail: chuntana.rea@mfu.ac.th
ONLINE MBISAND DEPRESSION IN COLLEGE AND UNIVERSITY STUDENTS 11
Downloaded by Univ Of Pittsburgh from www.liebertpub.com at 03/13/23. For personal use only.
