Psychiatry Research 326 (2023) 115322

Contents lists available at ScienceDirect

Psychiatry Research
journal homepage: www.elsevier.com/locate/psychres

Effects of resistance exercise training on depressive symptoms among

young adults: A randomized controlled trial
Darragh O’Sullivan a, b, *, Brett R. Gordon a, c, Mark Lyons a, d, Jacob D. Meyer e, Matthew
P. Herring a, b
a
Department of Physical Education and Sport Sciences, University of Limerick, Limerick, Ireland
b
Physical Activity for Health Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland
c
Public Health Sciences, Penn State College of Medicine, Pennsylvania, United States
d
Sport and Human Performance Research Centre, University of Limerick, Limerick, Ireland
e
Department of Kinesiology, Iowa State University, Iowa, United States

A R T I C L E I N F O A B S T R A C T

Keywords: Evidence supports the antidepressant effects of resistance exercise training (RET); however, findings among
Muscle strengthening exercise young adults at-risk for elevated depressive symptoms are limited. This randomized controlled trial examined the
Resistance training effects of eight weeks of ecologically-valid, guidelines-based RET, compared to a wait-list control, on depressive
Depression
symptoms among 55 young adults (26±5y; 36 female) with and without subclinical, or analogue, Generalized
Generalized anxiety disorder
Young adults
Anxiety Disorder (AGAD; Psychiatric Diagnostic Screening Questionnaire GAD subscale ≥6 and Penn State Worry
Questionnaire ≥45) and Major Depressive Disorder (AMDD). Following a three-week familiarization period,
participants completed one-on-one, twice-weekly RET sessions. The 16-item, self-reported Quick Inventory of
Depressive Symptomatology (QIDS) assessed depressive symptoms. RM-ANCOVAs examined between-group
differences, and significant interactions were decomposed with simple effects analysis. Hedges’ d effect sizes
(95%CI) quantified the magnitude of differences in change between groups across time. Stratified analyses were
conducted among subsamples with AMDD and AGAD. There were no baseline depressive symptom differences
between groups. Attendance was 83%, and compliance was 80%. RET induced statistically significant, clinically-
meaningful, large-magnitude reductions in depressive symptoms from baseline to week eight in the total (d =
1.01; [95%CI: 0.44–1.57]), AMDD (d = 1.71; [95%CI: 0.96–2.46]), and AGAD (d = 1.39; [95%CI: 0.55–2.24])
samples. These findings support guidelines-based RET as a promising treatment for mild depression.

1. Introduction depression treatments include medication and psychotherapy. The

effectiveness of antidepressant medications increase with greater
Depression has a lifetime prevalence of approximately 15.4% (Lim depression severity, but these medications have minimal effectiveness
et al., 2018). It is characterized by symptoms including depressed mood, for mild-to-moderate depressive symptoms (Cohen’s d = 0.11; number
loss of interest and/or pleasure, poor concentration, and disturbed sleep needed to treat (NNT)=16) (Fournier et al., 2010). Negative side-effects
or appetite (National Institute for Health and Care Excellence 2022). of antidepressant medications can also cause poor treatment adherence
Depression is associated with increased risk of suicide (Bernal et al., and low tolerability (Wang et al., 2018). Psychotherapy, while effective
2007), all-cause mortality (Cuijpers and Smit, 2002), cardiovascular (Cuijpers et al., 2020), can also be expensive and inaccessible. Conse­
disease (Correll et al., 2017), Alzheimer’s disease (Green et al., 2003), quently, alternative and augmentative treatments, such as exercise
social dysfunction (Fombonne et al., 2001), and a major economic training, are needed for depression.
burden, costing approximately €118 billion per annum in Europe The antidepressant effects of exercise are well-documented (Herring
(Sobocki et al., 2006). Approximately 55% of people with depression et al., 2012; Schuch et al., 2016), and are underpinned by several bio­
receive contact with primary care or specialist services, of which only logically plausible mechanisms. These mechanisms include improved
32% receive adequate treatment (Andrews et al., 2000). First-line regulation of the monoaminergic system (Dishman et al., 2006) and

* Corresponding author at: PG1-039, Department of Physical Education and Sport Sciences, University of Limerick, Limerick, Ireland.
E-mail address: [email protected] (D. O’Sullivan).

https://doi.org/10.1016/j.psychres.2023.115322
Received 9 March 2023; Received in revised form 20 June 2023; Accepted 23 June 2023
Available online 28 June 2023
0165-1781/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
D. O’Sullivan et al. Psychiatry Research 326 (2023) 115322

stress-response (Sheline et al., 2003; Duman, 2004), reductions in in­ 2.2. Participants
flammatory markers (Olson et al., 2007), and increased IGF-1 (Cassilhas
et al., 2010). The overwhelming majority of interventions that have Participants were recruited via posters, emails, and word of mouth,
examined the effects of exercise on depressive symptoms have used and initially completed an electronic battery of questionnaires to
aerobic exercise training (AET). The promising effects of resistance ex­ establish eligibility; Fig. 1 presents a flowchart of participant recruit­
ercise training (RET) on depressive symptoms remain understudied ment (Gordon et al., 2020a).
(Garber et al., 2011; Cooney et al., 2013; Netz, 2017). Recent Given that the primary outcome of these trials was AGAD symptoms,
meta-analytic evidence showed that RET significantly reduces depres­ participants were not recruited based on depressive symptoms. At
sive symptoms among otherwise healthy (mean effect size delta (Δ)= baseline, participants completed questionnaires including the 10-item
0.81) and mentally-ill adults (Δ=1.00) (Gordon et al., 2018). Only three GAD subscale of the Psychiatric Diagnostic Screening Questionnaire
trials focused on young adults (d = 0.52 to 1.48) (Doyne et al., 1987; (PDSQ-GAD) (Zimmerman and Mattia, 2001), 16-item Penn State Worry
Herring et al., 2011; Vizza et al., 2016), all of which included partici­ Questionnaire (PSWQ) (Meyer et al., 1990), and several other measures
pants with baseline depressive symptoms indicative of at least mild of signs and symptoms of GAD (e.g., depressive symptoms) (Rush et al.,
depression. One trial examined depressive symptom changes among 2003). Participants were classified as either AGAD (PDSQ-GAD ≥6 and
participants with clinically-meaningful anxiety (d = 0.52) (Herring PSWQ ≥45) or non-AGAD, and were diverted to parallel, identical RCTs
et al., 2011). This lack of research among young adults, and those with evaluating the effect of the same RET program on AGAD symptoms in
clinically-meaningful anxiety, are intertwined limitations of the avail­ young adults either with or without AGAD. Participants were then
able evidence. randomized, stratified by sex and AGAD status, to RET or a wait-list
Depressive disorders typically develop later in adulthood. However, control using www.randomizer.org. This study includes pooled data
the median age of onset has progressively declined (Kessler et al., 2003), from both RCTs (i.e., AGAD and non-AGAD).
and approximately 36.9% of cases first develop before 25 years of age Inclusion criteria for both RCTs were: i) age 18–40y; ii) no medical
(Solmi et al., 2022). Additionally, GAD typically develops in young contra-indication to participation in RET; and iii) no current pregnancy
adulthood (Baxter et al., 2013, 2014), and is highly comorbid with or lactation. Participants were not excluded for currently or previously
Major Depressive Disorder (MDD) (Lieb et al., 2005), suggesting that engaging in RET or physical activity at or before baseline. To account for
young adults with clinically-meaningful anxiety may be at-risk for RET-specific familiarity and baseline physical activity levels, partici­
elevated or clinically-meaningful depressive symptoms. The frequent pants were asked, in reference to the time at baseline assessment, how
comorbidity of depression and anxiety, and lack of evidence of the ef­ long they had been involved in a formalized RET program; this was
fects of RET, warrants a greater focus on these populations. determined in weeks and used to quantify participant training age.
Heterogeneity of the design and protocols of RET interventions is Although not an exclusion criterion, no participants were currently
another limitation that potentially limits the widespread prescription of involved in RET. RET and wait-list participants in both RCTs were
RET as a therapy for depression. The World Health Organization (WHO) advised to maintain their current levels of physical activity throughout
and American College of Sports Medicine (ACSM) muscle-strengthening the trial. Participants were not excluded for receiving other treatments
guidelines provide a widely implementable, ecologically-valid RET op­ for any mental health disorder. Ten participants were currently
tion. The WHO recommends engaging in muscle-strengthening activities receiving treatment for depression (pharmacotherapy-only (n = 8),
involving major muscle groups on two or more days per week (WHO, psychotherapy-only (n = 1), or both (n = 1)). No participants were
2011). More detailed guidelines from ACSM recommend progressive receiving treatment for GAD. Four participants currently receiving
strength training on two to three days per week, with 2–3 sets of 8–12 treatment completed the trial (RET n = 2, wait-list n = 2); the other six
repetitions for muscle strength benefits among individuals with limited withdrew (RET n = 2, wait-list n = 4).
or no RET experience (ACSM, 2019). RET designed in accordance with This trial was originally powered to detect changes in anxiety and
ACSM guidelines has demonstrated anxiolytic effects among young worry symptoms, which were the primary outcomes (Gordon et al.,
adults with (d = 0.71) (Gordon et al., 2021) and without (d = 0.85) 2020a). Based on previous meta-analytic evidence of the
(Gordon et al., 2020b) clinically-meaningful anxiety. small-to-moderate effect of RET on anxiety symptoms (Δ=0.31) (Gor­
Therefore, this secondary analysis of existing randomized controlled don et al., 2017), and the moderate-to-large effect of RET on depressive
trial (RCT) data extends past reports by quantifying the effects of symptoms (Δ=0.66) (Gordon et al., 2018), a priori power analyses were
ecologically-valid, guidelines-based RET on depressive symptoms conducted using G*Power 3.1. These analyses indicated that for each
among young adults with and without subclinical, or analogue, Gener­ parallel arm, sample sizes of 24 (12 per group) would provide >80%
alized Anxiety Disorder (AGAD) and Major Depressive Disorder statistical power (two-tailed α=0.05, four repeated measures) to detect a
(AMDD). small-to-moderate effect of RET on anxiety symptoms, and a slightly
lower than anticipated (i.e., more conservative) small-to-moderate ef­
2. Methods fect of RET on depressive symptoms.

This pilot efficacy trial adhered to the Consolidated Standards of 2.3. RET intervention
Reporting Trials (CONSORT) Checklist (Schulz et al., 2010).
RET was designed in accordance with WHO and ACSM guidelines
(WHO, 2011; ACSM, 2019). The eight-week, twice-weekly intervention
2.1. Trial design increased resistance progressively, such that the participant could
complete two sets of between 8 and 12 repetitions before a deterioration
Secondary analyses of existing data from two parallel, eight-week in lifting form, or failure to complete a repetition. The investigator
RCTs (ClinicalTrials.gov Identifier: NCT04116944) were completed; specified the resistance in accordance with guidelines rather than
the full methods (Gordon et al., 2020a) and primary outcomes (Gordon self-selection by participants. Load progressions were small and gradual.
et al., 2020b, 2021) are published. The research protocol was approved When participants completed two sets of 12 repetitions on an exercise,
by the University’s Research Ethics Committee (EHSREC No: load was increased by 5% in the following session. When participants
2017_03_18_EHS); all participants provided written informed consent failed to complete two sets of at least 8 repetitions on an exercise, load
prior to participation which was not compensated. This trial had rolling was decreased by 5% in the following session. The eight exercises were
recruitment; data collection commenced in January 2018, and barbell back squat, barbell bench press, hexagon bar deadlift, seated
concluded in June 2019. dumbbell shoulder lateral raise, barbell bent over rows, dumbbell

2
D. O’Sullivan et al. Psychiatry Research 326 (2023) 115322

Fig. 1. flowchart of participant recruitment.

lunges, seated dumbbell curls, and abdominal crunches. After baseline 2.6. Covariates
assessment, participants randomized to RET completed a three-week,
twice-weekly familiarization process to ensure safety, correct lifting Baseline physical activity was assessed using an online, self-report
technique, and that the intervention was delivered at the correct resis­ version of a seven-day Physical Activity Recall (Blair et al., 1985).
tance, starting at week one. Exercise sessions lasted approximately 25 Participants reported time engaged in sleep, and moderate, hard, and
min and were fully supervised on a one-to-one basis in a small, private, very hard activities during the prior week. Estimated energy expenditure
university-owned RET facility. All investigators were fully trained in was calculated as kilocalories per week. According to thresholds vali­
delivering the exercise protocol consistently and identifying proper and dated by Dishman and Steinhardt (1988) among young adult university
improper lifting mechanics. Prior to RET sessions, participants samples, participants can be considered inactive (<245 kcal⋅kg− 1⋅wk− 1)
completed primary and secondary outcome questionnaires in the RET to highly active (≥280 kcal⋅kg− 1⋅wk− 1).
facility. Further specifics of the RET intervention have been published
previously (Gordon et al., 2020a). 2.7. Intervention fidelity and manipulation check

2.4. Control condition To calculate attendance percentage, the number of RET sessions
attended was divided by 16 (two sessions per week for eight weeks) and
Participants randomized to the wait-list condition completed online multiplied by 100. To calculate compliance to the RET protocol, the
questionnaires weekly, and were subsequently offered the RET inter­ number of sets in which at least eight repetitions were completed was
vention upon completion of their wait-list condition, but no data were divided by the total number of sets prescribed across the eight weeks (n
collected. = 256) and multiplied by 100.
To facilitate setting of the initial load, and to quantify anticipated
2.5. Primary outcome changes in strength as a manipulation check, participants completed a
five-repetition maximum (5RM) assessment for the barbell bench press,
The primary outcome for this analysis was depressive symptom barbell back squat, and hexagon bar deadlift at baseline and post-
severity among the total sample. The 16-item self-report version of the intervention. Total strength changes were measured as load changes
Quick Inventory of Depressive Symptomatology (QIDS) (Rush et al., on all three 5RM lifts combined. During the six familiarization sessions,
2003) was used to assess depressive symptom severity at baseline, week participants completed two familiarizations with the 5RM process, and
one, four, and eight; symptoms were measured before the RET sessions. one maximal 5RM assessment.
The QIDS has previously shown strong internal consistency (a = 0.86),
and is a treatment-sensitive measure of depressive symptom severity 2.8. Statistical analyses
(Trivedi et al., 2004). The internal consistency of the QIDS for the total
sample was a = 0.77 (95%CI: 0.66 to 0.85). Correlations between Data analyses were performed using SPSS 26.0. Four percent of data
repeated measures at baseline and week one were r = 0.67 (p<0.001) were missing from the QIDS. Missing data were imputed for nine par­
and r = 0.87 (p<0.001) for the RET and wait-list groups, respectively. ticipants: sex and time-variant responses for each variable were entered

3
D. O’Sullivan et al. Psychiatry Research 326 (2023) 115322

as predictors into separate multiple linear regression models for condi­ 3.1. Intervention fidelity and manipulation check
tion, and predicted values were retained. Participants (n = 44) were
excluded if they were missing primary outcome data for more than one The average attendance to the RET intervention was 83% (13 out of
time-point. These participants were predominately those who withdrew 16 sessions). The average compliance with RET was 80% (205 out of 256
following randomization, but prior to completing any exercise or wait- sets). As previously reported, these attendance and compliance rates
list control condition protocol. Intention-to-treat analyses included all indicated that participants missed approximately three sessions over the
participants, and analyses of complete-cases only are reported as course of the eight-week intervention, and were almost fully compliant
sensitivity analyses. when they attended. No adverse events arose from trial participation;
Independent samples t-tests examined baseline differences between however, one participant randomized to RET reported a headache dur­
conditions and sexes. The magnitudes of baseline differences were ing an exercise bout. The investigator supervising the bout immediately
quantified using Cohen’s d effect sizes (Rosenthal et al., 1994). A stopped the session, and the participant subsequently withdrew from the
two-group (RET/wait-list) x four time-point (baseline/week one/week trial after consulting with a physician. The average 6–20 rating of
four/week eight) repeated measures ANCOVA examined differences perceived exertion was 14±1 (i.e., between somewhat hard and hard),
between RET and wait-list; age, sex, and baseline physical activity were and the average muscle soreness was 4 ± 2 (i.e., between mild and some
covariates. A two-sex x two-group x four time-point repeated measures soreness present) (Heath, 1998). As anticipated, participants in the RET
ANCOVA, adjusted for baseline physical activity, examined potential intervention significantly increased their total strength (t(20)=− 9.2,
sex-related differences in the effects of RET compared to wait-list. The p<0.001, Cohen’s d = 2.01[95%CI: 1.25 to 2.75]), mean increase: 23.4%
Huynh-Feldt adjustment was applied when the assumption of sphericity ±13.2).
was violated. Significant interactions were decomposed using simple
effects analyses. Standardized mean differences (SMD) quantified the 3.2. Depressive symptoms among the total sample
magnitude of within-condition change. The magnitude of difference in
outcome change between groups was quantified by Hedges’ d effect sizes Table 2 presents descriptives, SMDs, and Hedges’ d [95%CI] effect
by subtracting the mean change in the wait-list from the mean change in sizes for changes in depressive symptoms from baseline. Results for
the RET condition and dividing this difference by the pooled standard males and females are reported in Table 1 in the Supplementary
deviation of baseline scores (Hedges and Olkin, 1985); effect sizes were Materials.
adjusted for small sample size bias and calculated such that larger At baseline, depressive symptom scores for the total sample were
depressive symptom reductions among RET compared to wait-list indicative of mild depression (QIDS: RET: 8.7; wait-list: 8.0; 6–10=mild
resulted in positive effect sizes. Changes in strength were examined depression). Among the total sample, no significant sex X group X time
using paired-samples t-tests. Associations between changes in strength interaction was found for depressive symptoms (F(2.3,114)=0.942, p =
and changes in depressive symptoms were quantified using Pearson 0.403). A significant group X time interaction was found
correlation coefficients of associations between change scores. (F(2.3,113)=10.18, p<0.001, d = 1.01[0.44 to 1.57]), such that RET
Sub-analyses were performed for four subsamples: young adults with significantly reduced depressive symptoms from baseline to week eight
AGAD, without AGAD, with AMDD (QIDS ≥ 6), and without AMDD compared to the wait-list control (mean difference [Mdiff]=− 5.82,
(QIDS ≤ 5). p<0.001). When examining raw changes, depressive symptoms reduced
among 39 participants (RET=24; wait-list=15) from baseline to week
3. Results eight; symptoms increased (i.e., worsened) among eight participants
(RET=2; wait-list=6), of which three transitioned from non-AMDD to
Tables 1 and 2 present baseline participant characteristics and AMDD (RET=2; wait-list=1), and did not change among eight wait-list
changes in depressive symptoms across the trial, respectively, for the participants. Findings were materially the same for intention-to-treat
total sample and each subsample. There were no baseline differences analyses (d = 0.88[0.45 to 1.31]) and analyses of complete cases only
between groups for any outcomes, supporting successful randomization; (d = 0.89[0.28 to 1.50]). Changes in strength were not significantly
there were also no baseline differences between sexes for any outcomes associated with changes in depressive symptoms (r(21)=− 0.12, p ≥
(all p>0.14). 0.62).

3.3. Depressive symptoms and remission among AMDD, AGAD, and non-
AGAD subsamples
Table 1
Total sample participant characteristics and baseline differences between Results for dichotomized subsamples (e.g., AMDD without AGAD)
groups. are reported in Table 2 in the supplementary materials. As would be
RET (n WL (n expected, given the comorbidity of GAD and MDD, there is comorbidity
= 26) = 29) in the current sample such that some participants were classified (and
Variables Mean Mean t p Cohen’s
thus included in analyses) for both AGAD and AMDD.
(SD) (SD) d (95%CI)
Baseline depressive symptoms among the AMDD and AGAD samples
% Female 65.4 62.1 were indicative of moderate depression (all QIDS>10). Among the
Age (y) 25.8 27.5 − 1.1 0.271 − 0.30
(5.7) (5.7) (− 0.83 to
AMDD sample, a significant group X time interaction was found for
0.23) depressive symptoms (F(2.5,80)=10.72, p<0.001, d = 1.71[0.96 to 2.46]).
BMI 24.6 24.5 0.1 0.893 0.02 (− 0.51 RET significantly reduced depressive symptoms from baseline to week
(4.1) (4.1) to 0.55) eight compared to the wait-list control ([Mdiff]=− 7.73, p<0.001). RET
Physical Activity 263.2 271.1 − 0.8 0.442 − 0.21
reduced depressive symptoms among every member of the AMDD
(kcal⋅kg¡1⋅wk¡1) (30.2) (43.9) (− 0.74 to
0.32) sample; 18 of the 19 (95%) RET participants remitted (QIDS<6) from
Depressive Symptoms 8.7 (4.2) 8.0 0.6 0.578 0.15 (− 0.38 AMDD by week eight. Among the wait-list (n = 18), 4 of 18 participants
(QIDS) (5.3) to 0.68) (22%) remitted from AMDD by week eight; symptoms remained the
Physical activity was assessed via 7-day Physical Activity Recall. same among four, increased among four, and reduced without remission
RET=Resistance Exercise Training; WL=Wait-list; SD=Standard Deviation; among six (33%).
y=Years; BMI=Body Mass Index; QIDS=Quick Inventory of Depressive Significant group X time interactions were also found for depressive
Symptomatology. symptoms among young adults with (F(2.2,49)=5.81, p ≤ 0.004, d = 1.39

4
D. O’Sullivan et al. Psychiatry Research 326 (2023) 115322

Table 2
Changes in depressive symptoms from baseline.
Sample Group n Baseline Week 1 SMD Hedges’ d from Week 4 SMD Hedges’ d from Week 8 SMD Hedges’ d from
baseline baseline baseline

Total RET 26 8.7 (4.2) 5.2 (3.4) 0.92 0.59 (0.05 to 1.13)* 4.5 (3.0) 1.15 0.66 (0.13 to 1.22)* 2.8 1.79 1.01 (0.44 to 1.57)*
(2.0)
WL 29 8.0 (5.3) 7.4 (5.5) 0.11 7.1 (6.0) 0.16 7.0 0.19
(5.5)
AMDD RET 19 10.6 6.2 (3.4) 1.31 0.90 (0.22 to 1.57)* 4.9 (2.8) 1.86 1.14 (0.45 to 1.84)* 2.7 3.01 1.71 (0.96 to 2.46)*
(3.3) (1.7)
WL 18 11.3 10.2 0.25 9.8 (6.1) 0.29 9.7 0.35
(3.9) (5.0) (5.2)
AGAD RET 12 11.4 7.1 (3.6) 1.15 0.79 (0.00 to 1.58)* 5.8 (2.7) 1.67 1.07 (0.26 to 1.88)* 3.3 2.64 1.39 (0.54 to 2.23)*
(3.9) (1.9)
WL 15 11.3 10.7 0.12 10.7 0.11 9.7 0.29
(5.0) (5.2) (6.3) (6.1)

AMDD=Analogue Major Depressive Disorder; AGAD=Analogue Generalized Anxiety Disorder; n=sample size; SMD=Standardized Mean Difference; RET=Resistance
Exercise Training; WL=Wait-list; *indicates a significant difference from baseline score in simple effects analysis.

[0.54 to 2.23]) and without (F(2.7,62)=5.46, p = 0.003, d = 1.24[0.43 to the same among males (d = 1.11) and females (d = 0.89). Based on the
2.05]) AGAD. RET significantly reduced depressive symptoms from threshold for clinical meaningfulness of at least half a standard deviation
baseline to week eight among those with ([Mdiff]=− 7.81, p<0.001) and reduction from baseline scores (Norman et al., 2003), the
without ([Mdiff]=− 4.13, p<0.001) AGAD compared to the wait-list large-magnitude reduction in depressive symptoms from baseline to
control. Of the 27 participants with AGAD, 24 had comorbid AMDD. week eight (d = 1.01[0.44 to 1.57]) is clinically meaningful. This anti­
Table 3 presents Hedges’ d effect sizes (95%CIs) for changes in depressant effect of RET is larger than previously reported findings for
depressive symptoms across successive measurement time-points. aerobic and resistance exercise (Δ=0.62) (Cooney et al., 2013), and for
RET among older (≥55 years) adults (Δ=0.72) (Gordon et al., 2018),
4. Discussion and comparable to that for mentally-ill adults (Δ=1.00) (Gordon et al.,
2018). It is also larger than antidepressant effects of RET found in two
WHO and ACSM guidelines-based RET was an efficacious interven­ previous experimental studies of young adults (d = 0.52, 0.84) (Herring
tion for reducing depressive symptoms among young adults with limited et al., 2011; Vizza et al., 2016), but smaller than effects found in another
RET experience. The efficacy of RET also appeared to be high among trial of clinically depressed young adults (Δ=1.48) (Doyne et al., 1987).
young adults with AGAD and AMDD, supporting RET as a promising This is the third RET trial to report large-magnitude reductions in
treatment for mild or subclinical depression among young adults. depressive symptoms among young adults, and the first to do so using
ecologically-valid RET designed in accordance with WHO and ACSM
4.1. RET among the total sample guidelines.
The large-magnitude antidepressant effect found in this study sup­
Attendance of 83% and compliance of 80%, and the absence of ports RET as a promising treatment for mild or subclinical depression
adverse events occurring from RET, supports that the RET intervention, among young adults. This antidepressant effect is larger than those
which was representative of common, real-world RET practises, was found for antidepressant medications, which have shown limited
feasible and tolerable. Further, a beneficial side-effect of RET was the effectiveness for reducing mild-to-moderate depressive symptoms
large-magnitude increase in strength (d = 2.01); however, changes in (Cohen’s d = 0.11) (Fournier et al., 2010), and are unlikely to be used for
strength were not associated with changes in depressive symptoms, and, preventing the onset of clinical depression. The increased likelihood of
consequently, RET that is programmed to improve strength may not be young people who endorse depressed mood later developing clinical
required for antidepressant benefits from RET among young adults with depression (odds ratio=2.52) (Wolitzky-Taylor et al., 2014) increases
limited RET experience. Feasible, tolerable, and conservatively pro­ the potential importance of this finding. RET may be a particularly
gressive RET that facilitates continuous engagement with RET activity effective intervention for preventing the onset of clinical depression
may induce antidepressant benefits among young adults with limited among individuals lower on the depressive symptom severity spectrum,
RET experience. while concurrently improving aspects of physical health and function.
Compared to a wait-list control group, RET significantly reduced
depressive symptoms from baseline to week one (i.e., pre-post three- 4.2. RET among the AMDD and AGAD samples
week familiarization protocol), and week four to week eight. Although
reductions in depressive symptoms plateaued from week one to week The prevalent comorbidity of AMDD and AGAD in the current sam­
four, the antidepressant effect of RET from baseline to week eight ple reflects the frequent comorbidity of GAD and MDD (Lieb et al.,
appeared larger than that of baseline to week four, and was materially 2005). The current study is the first to investigate the antidepressant
effects of ecologically valid, guidelines-based RET among young adults
Table 3 with clinically-meaningful symptoms of GAD, in whom depressive
Effect of RET on depressive symptoms across successive time-points. symptoms are frequently comorbid and may impact overall symptom
Sample Hedges’ d Hedges’ d Hedges’ d severity, treatment compliance, and treatment response. RET induced
Baseline to Week 1 Week 1 to Week 4 Week 4 to Week 8 large-magnitude, clinically-meaningful reductions in depressive symp­
Total 0.59 (0.05 to 1.13)* 0.09 (− 0.44 to 0.61) 0.33 (− 0.21 to 0.86)* toms from baseline to week eight among the AMDD and AGAD samples.
AMDD 0.90 (0.22 to 1.57)* 0.21 (− 0.44 to 0.85) 0.44 (− 0.22 to 1.09)* Depressive symptoms among the AMDD and AGAD samples were similar
AGAD 0.79 (0.00 to 1.58)* 0.28 (− 0.49 to 1.04) 0.29 (− 0.47 to 1.05)* to that of the total sample at week eight (QIDS: AMDD: 2.7; AGAD: 3.3;
AMDD: Analogue Major Depressive Disorder; AGAD=Analogue Generalized total: 2.8), and almost the entire AMDD sample (95%) remitted from
Anxiety Disorder;. AMDD. Consequently, RET reduced depressive symptoms among the
*
Indicates a significant difference from previous time-point in simple effects AMDD and AGAD samples to the degree whereby those initially meeting
analysis. criteria for AMDD no longer met those criteria by week eight.

5
D. O’Sullivan et al. Psychiatry Research 326 (2023) 115322

The antidepressant effect of RET among the AMDD sample in this and mediation analyses to examine such response variation. Using the
study is larger in magnitude than that of previous meta-analytic evi­ QIDS as the diagnostic measure of depression status for the AMDD
dence for RET among participants with symptoms indicative of mild-to- sample was also a potential limitation. The QIDS is primarily used to
moderate depression (Gordon et al., 2018). This antidepressant effect is measure changes in symptoms across time, rather than diagnosing
also larger than those found in previous RCTs of young adults with major depression status. Future trials should use diagnostic questionnaires to
or minor depression (Doyne et al., 1987), and depressive symptoms diagnose depression status. The effect of RET on depressive symptoms
indicative of mild-to-severe depression (Herring et al., 2011; Vizza et al., among young adults with clinically-meaningful MDD with or without
2016). There are now four RCTs that have examined the effect of RET on comorbid, clinically-meaningful GAD, remains unclear. Given the
depressive symptoms among young adults with either prevalence and additional burden of comorbid GAD and MDD, the
clinically-diagnosed depression, or symptoms indicative of at least mild associated increased treatment difficulty, and the potential synergistic
depression. Three RCTs found large antidepressant effects (d = 0.84 to effect, whereby simultaneous reductions in depressive and anxiety
1.71) using full-body RET among young adults with symptoms indica­ symptoms following RET augment each other, continued evaluation of
tive of moderate-to-severe depression, while one found a moderate an­ RET as a treatment for comorbid GAD and MDD is warranted.
tidepressant effect (d = 0.52) using lower body RET among young adults
with symptoms indicative of mild depression. Consequently, higher 5. Conclusion
baseline depressive symptoms, using full-body RET, or both, potentially
increase the effectiveness of RET for reducing depressive symptoms Eight weeks of ecologically-valid RET designed in accordance with
among young adults. WHO and ACSM guidelines resulted in clinically-meaningful, large-
The antidepressant effect found among the AGAD sample is larger in magnitude reductions in depressive symptoms among an otherwise
magnitude than that of the only previous experimental trial to examine healthy sample of young adults. Sub-analyses also revealed large anti­
the effects of RET on depressive symptoms among young adults with depressant effects among participants with AMDD and AGAD. There is a
clinically-meaningful anxiety (i.e., GAD) (Herring et al., 2011). Co­ potential synergistic effect among those with AGAD, such that re­
morbidity is a fundamental characteristic of GAD, which is associated ductions in depressive and anxiety symptoms following RET augment
with an increased risk of depression (odds ratio=8.7) (Copeland et al., each other. The large-magnitude increase in strength was a beneficial
2014). Such comorbidity among mental health disorders intensifies side-effect of RET, and was not associated with changes in depressive
symptom severity, and makes treating both the primary and secondary symptoms.
disorder more difficult. The increased difficulty of treating comorbid Research Support
mental health disorders adds significance to the reductions in both Brett R. Gordon was funded by the Irish Research Council under the
depressive and anxiety (d = 0.71) (Gordon et al., 2021) symptoms found Government of Ireland Postgraduate Programme.
among the AGAD sample in this RET trial. The large magnitude anti­ Relationships
depressant effect of RET among those with AGAD suggests that RET may There are no additional relationships to disclose.
be a promising primary or adjunctive treatment for moderate depressive Patents and Intellectual Property
symptoms among young adults with clinically-meaningful anxiety, and There are no patents to disclose.
little previous RET experience. The size of the large magnitude re­ Other Activities
ductions in depressive symptoms among participants with AGAD (d = There are no additional activities to disclose.
1.39) could be due to their higher baseline depressive symptoms (QIDS:
11.4) (Gordon et al., 2018). There could also be a synergistic effect, Funding
whereby the simultaneous reductions in symptoms of GAD and
depression augment each other, resulting in further symptom re­ B.G. was funded for this research by a postgraduate scholarship from
ductions. The overlapping aetiologies of GAD and MDD may explain the Irish Research Council. B.G. is currently supported by a National
these simultaneous antidepressant and anxiolytic effects. Greater psy­ Institute on Aging F32 grant F32AG078229.
chosocial and psychobiological benefits of RET among those with a
greater anxiety symptom burden may have also contributed to the CRediT authorship contribution statement
larger-magnitude antidepressant effect among the AGAD sample.
Darragh O’Sullivan: Conceptualization, Formal analysis, Writing –
4.3. Limitations and future research original draft, Writing – review & editing, Visualization. Brett R. Gor­
don: Conceptualization, Methodology, Formal analysis, Investigation,
Limitations of this study include the lack of an attention-control Writing – original draft, Writing – review & editing, Visualization, Su­
condition. Social interaction, attention, mastery experiences, and ex­ pervision. Mark Lyons: Conceptualization, Methodology, Writing – re­
pectations for improvement, all of which cannot be controlled for with a view & editing, Visualization, Supervision. Jacob D. Meyer: Writing –
non-active control group, may have contributed to the antidepressant review & editing, Visualization. Matthew P. Herring: Conceptualiza­
effect found among the RET group. Still, the comparison to a non-active tion, Methodology, Writing – review & editing, Visualization,
control gives a baseline indication of the efficacy of a RET intervention Supervision.
for reducing depressive symptoms. Future trials should use active
attention-controls, such as minimal-intensity exercise, to examine the Declaration of Competing Interest
antidepressant effect of RET while controlling for non-exercise inter­
vention components. Further, there were no post-intervention follow- Declarations of interest: none.
ups in the current study; given the commonality of relapse following the
cessation of treatment among patients with MDD (Ramana et al., 1995), Supplementary materials
long-term follow-ups are needed to examine whether depressive symp­
tom reductions, and involvement in RET, are maintained. Future trials Supplementary material associated with this article can be found, in
should also control for the use of contraceptive medications, due to their the online version, at doi:10.1016/j.psychres.2023.115322.
effects on reproductive hormonal balance. Inter-individual variation in
depressive symptom responses to RET also remains unclear. As this was
a pilot efficacy trial, though adequately powered, the sample size was
small; future trials with larger sample sizes will allow for moderation

6
D. O’Sullivan et al. Psychiatry Research 326 (2023) 115322

References Heath, E.M., 1998. Borg’s perceived exertion and pain scales. Med. Sci. Sports Exerc. 30,
1461.
Hedges, L.V., Olkin, I., 1985. Statistical Methods For Meta-Analysis. Academic Press,
American College of Sports Medicine, 2019. ACSM guidelines for strength training.
New York.
https://www.acsm.org/blog-detail/acsm-certified-blog/2019/07/31/acsm-guidel
Herring, M.P., Jacob, M.L., Suveg, C., O’Connor, P.J., 2011. Effects of short-term exercise
ines-for-strength-training-featured-download (accessed 11 Mar 2022).
training on signs and symptoms of generalized anxiety disorder. Ment. Health Phys.
Andrews, G., Sanderson, K., Corry, J., Lapsley, H.M., 2000. Using epidemiological data to
Act. 4 (2), 71–77.
model efficiency in reducing the burden of depression. J. Ment. Health Policy Econ. 3
Herring, M.P., Puetz, T.W., O’Connor, P.J., Dishman, R.K., 2012. Effect of exercise
(4), 175–186.
training on depressive symptoms among patients with a chronic illness: a systematic
Baxter, A.J., Scott, K.M., Ferrari, A.J., Norman, R.E., Vos, T., Whiteford, H.A., 2014.
review and meta-analysis of randomized controlled trials. Arch. Intern. Med. 172 (2),
Challenging the myth of an “epidemic” of common mental disorders: trends in the
101–111.
global prevalence of anxiety and depression between 1990 and 2010. Depress.
Kessler, R.C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K.R., Rush, A.J.,
Anxiety. 31 (6), 506–516.
Walters, E.E., Wang, P.S., 2003. The epidemiology of major depressive disorder:
Baxter, A.J., Scott, K.M., Vos, T., Whiteford, H.A., 2013. Global prevalence of anxiety
results from the National Comorbidity Survey Replication (NCS-R). JAMA 289 (23),
disorders: a systematic review and meta-regression. Psychol. Med. 43 (5), 897–910.
3095–3105.
Bernal, M., Haro, J.M., Bernert, S., Brugha, T., de Graaf, R., Bruffaerts, R., Lépine, J.P., de
Lieb, R., Becker, E., Altamura, C., 2005. The epidemiology of generalized anxiety
Girolamo, G., Vilagut, G., Gasquet, I., 2007. Risk factors for suicidality in Europe:
disorder in Europe. Eur. Neuropsychopharmacol. 15 (4), 445–452.
results from the ESEMED study. J. Affect. Disord. 101 (1–3), 27–34.
Lim, G.Y., Tam, W.W., Lu, Y., Ho, C.S., Zhang, M.W., Ho, R.C., 2018. Prevalence of
Blair, S.N., Haskell, W.L., Ho, P., Paffenbarger Jr, R.S., Vranizan, K.M., Farquhar, J.W.,
depression in the community from 30 countries between 1994 and 2014. Sci. Rep. 8
Wood, P.D., 1985. Assessment of habitual physical activity by a seven-day recall in a
(1), 1–10.
community survey and controlled experiments. Am. J. Epidemiol. 122 (5), 794–804.
Meyer, T.J., Miller, M.L., Metzger, R.L., Borkovec, T.D., 1990. Development and
Cassilhas, R.C., Antunes, H.K.M., Tufik, S., De Mello, M.T., 2010. Mood, anxiety, and
validation of the Penn State Worry Questionnaire. Behav. Res. Ther. 28 (6),
serum IGF-1 in elderly men given 24 weeks of high resistance exercise. Percept. Mot.
487–495.
Skills. 110 (1), 265–276.
National Institute for Health and Care Excellence, 2022. Depression in adults: Treatment
Cooney, G.M., Dwan, K., Greig, C.A., Lawlor, D.A., Rimer, J., Waugh, F.R., McMurdo, M.,
and Management. https://www.nice.org.uk/guidance/ng222 (Accessed August 2).
Mead, G.E., 2013. Exercise for depression. Cochrane Database Syst. Rev. (9).
Netz, Y., 2017. Is the comparison between exercise and pharmacologic treatment of
Copeland, W.E., Angold, A., Shanahan, L., Costello, E.J., 2014. Longitudinal patterns of
depression in the clinical practice guideline of the American College of Physicians
anxiety from childhood to adulthood: the Great Smoky Mountains Study. J. Am.
evidence-based? Front. Pharmacol. 8, 257.
Acad. Child Adolesc. Psychiatry. 53 (1), 21–33.
Norman, G.R., Sloan, J.A., Wyrwich, K.W., 2003. Interpretation of changes in health-
Correll, C.U., Solmi, M., Veronese, N., Bortolato, B., Rosson, S., Santonastaso, P., Thapa-
related quality of life: the remarkable universality of half a standard deviation. Med.
Chhetri, N., Fornaro, M., Gallicchio, D., Collantoni, E., 2017. Prevalence, incidence
Care. 41 (5), 582–592.
and mortality from cardiovascular disease in patients with pooled and specific severe
Olson, T.P., Dengel, D., Leon, A., Schmitz, K., 2007. Changes in inflammatory biomarkers
mental illness: a large-scale meta-analysis of 3,211,768 patients and 113,383,368
following one-year of moderate resistance training in overweight women. Int. J.
controls. World Psychiatry 16 (2), 163–180.
Obes. 31 (6), 996–1003.
Cuijpers, P., Karyotaki, E., Eckshtain, D., Ng, M.Y., Corteselli, K.A., Noma, H., Quero, S.,
Ramana, R., Paykel, E., Cooper, Z., Hayhurst, H., Saxty, M., Surtees, P., 1995. Remission
Weisz, J.R., 2020. Psychotherapy for depression across different age groups: a
and relapse in major depression: a two-year prospective follow-up study. Psychol.
systematic review and meta-analysis. JAMA Psychiatry 77 (7), 694–702.
Med. 25 (6), 1161–1170.
Cuijpers, P., Smit, F., 2002. Excess mortality in depression: a meta-analysis of community
Rosenthal, R., Cooper, H., Hedges, L., 1994. Parametric measures of effect size. In:
studies. J. Affect. Disord. 72 (3), 227–236.
Cooper, H., Hedges, L. (Eds.), The Handbook of Research Synthesis. Russell Sage
Dishman, R.K., Berthoud, H.R., Booth, F.W., Cotman, C.W., Edgerton, V.R., Fleshner, M.
Foundation, pp. 231–244.
R., Gandevia, S.C., Gomez-Pinilla, F., Greenwood, B.N., Hillman, C.H., 2006.
Rush, A.J., Trivedi, M.H., Ibrahim, H.M., Carmody, T.J., Arnow, B., Klein, D.N.,
Neurobiology of exercise. Obesity. 14 (3), 345–356.
Markowitz, J.C., Ninan, P.T., Kornstein, S., Manber, R., 2003. The 16-Item Quick
Dishman, R.K., Steinhardt, M., 1988. Reliability and concurrent validity for a 7-d re-call
Inventory of Depressive Symptomatology (QIDS), clinician rating (QIDS-C), and self-
of physical activity in college students. Med. Sci. Sports Exerc. 20 (1), 14–25.
report (QIDS-SR): a psychometric evaluation in patients with chronic major
Doyne, E.J., Ossip-Klein, D.J., Bowman, E.D., Osborn, K.M., McDougall-Wilson, I.B.,
depression. Biol. Psychiatry. 54 (5), 573–583.
Neimeyer, R.A., 1987. Running versus weight lifting in the treatment of depression.
Schuch, F.B., Vancampfort, D., Richards, J., Rosenbaum, S., Ward, P.B., Stubbs, B., 2016.
J. Consult. Clin. Psychol. 55 (5), 748–754.
Exercise as a treatment for depression: a meta-analysis adjusting for publication bias.
Duman, R.S., 2004. Role of neurotrophic factors in the etiology and treatment of mood
J. Psychiatr. Res. 77, 42–51.
disorders. Neuromolecular Med 5 (1), 11–25.
Schulz, K.F., Altman, D.G., Moher, D., 2010. CONSORT 2010 statement: updated
Fombonne, E., Wostear, G., Cooper, V., Harrington, R., Rutter, M., 2001. The Maudsley
guidelines for reporting parallel group randomised trials. BMC Med 11 (1), 1–8.
long-term follow-up of child and adolescent depression: 2. Suicidality, criminality
Sheline, Y.I., Gado, M.H., Kraemer, H.C., 2003. Untreated depression and hippocampal
and social dysfunction in adulthood. Br. J. Psychiatry. 179 (3), 218–223.
volume loss. Am. J. Psychiatry. 160 (8), 1516–1518.
Fournier, J.C., DeRubeis, R.J., Hollon, S.D., Dimidjian, S., Amsterdam, J.D., Shelton, R.
Sobocki, P., Jönsson, B., Angst, J., Rehnberg, C., 2006. Cost of depression in Europe.
C., Fawcett, J., 2010. Antidepressant drug effects and depression severity: a patient-
J. Ment. Health Policy Econ. 9 (2), 87–98.
level meta-analysis. JAMA 303 (1), 47–53.
Solmi, M., Radua, J., Olivola, M., Croce, E., Soardo, L., Salazar de Pablo, G., Il Shin, J.,
Garber, C.E., Blissmer, B., Deschenes, M.R., Franklin, B.A., Lamonte, M.J., Lee, I.-M.,
Kirkbride, J.B., Jones, P., Kim, J.H., 2022. Age at onset of mental disorders
Nieman, D.C., Swain, D.P., 2011. American College of Sports Medicine position
worldwide: large-scale meta-analysis of 192 epidemiological studies. Mol.
stand. Quantity and quality of exercise for developing and maintaining
Psychiatry. 27 (1), 281–295.
cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy
Trivedi, M.H., Rush, A., Ibrahim, H., Carmody, T., Biggs, M., Suppes, T., Crismon, M.,
adults: guidance for prescribing exercise. Med. Sci. Sports Exerc. 43 (7), 1334–1359.
Shores-Wilson, K., Toprac, M., Dennehy, E., 2004. The Inventory of Depressive
Gordon, B.R., McDowell, C.P., Hallgren, M., Meyer, J.D., Lyons, M., Herring, M.P., 2018.
Symptomatology, Clinician Rating (IDS-C) and Self-Report (IDS-SR), and the Quick
Association of efficacy of resistance exercise training with depressive symptoms:
Inventory of Depressive Symptomatology, Clinician Rating (QIDS-C) and Self-Report
meta-analysis and meta-regression analysis of randomized clinical trials. JAMA
(QIDS-SR) in public sector patients with mood disorders: a psychometric evaluation.
Psychiatry 75 (6), 566–576.
Psychol. Med. 34 (1), 73–82.
Gordon, B.R., McDowell, C.P., Lyons, M., Herring, M.P., 2017. The effects of resistance
Vizza, L., Smith, C.A., Swaraj, S., Agho, K., Cheema, B.S., 2016. The feasibility of
exercise training on anxiety: a meta-analysis and meta-regression analysis of
progressive resistance training in women with polycystic ovary syndrome: a pilot
randomized controlled trials. Sports Med 47 (12), 2521–2532.
randomized controlled trial. BMC Sports Sci. Med. Rehabil. 8 (1), 1–12.
Gordon, B.R., McDowell, C.P., Lyons, M., Herring, M.P., 2020a. Acute and chronic effects
Wang, S., Han, C., Bahk, W., Lee, S., Patkar, A.A., Masand, P.S., Pae, C., 2018. Addressing
of resistance exercise training among young adults with and without analogue
the side effects of contemporary antidepressant drugs: a comprehensive review.
generalized anxiety disorder: a protocol for pilot randomized controlled trials. Ment.
Chonnam Med. J. 54 (2), 101–112.
Health Phys. Act. 18, 100321.
Wolitzky-Taylor, K., Dour, H., Zinbarg, R., Mineka, S., Vrshek-Schallhorn, S., Epstein, A.,
Gordon, B.R., McDowell, C.P., Lyons, M., Herring, M.P., 2020b. Resistance exercise
Bobova, L., Griffith, J., Waters, A., Nazarian, M., 2014. Experiencing core symptoms
training for anxiety and worry symptoms among young adults: a randomized
of anxiety and unipolar mood disorders in late adolescence predicts disorder onset in
controlled trial. Sci. Rep. 10 (1), 1–9.
early adulthood. Depress. Anxiety. 31 (3), 207–213.
Gordon, B.R., McDowell, C.P., Lyons, M., Herring, M.P., 2021. Resistance exercise
World Health Organization, 2011. Global recommendations on physical activity for
training among young adults with analogue generalized anxiety disorder. J. Affect.
health. https://www.who.int/dietphysicalactivity/global-PA-recs-2010.pdf
Disord. 281, 153–159.
(accessed 11 Mar 2022).
Green, R.C., Cupples, L.A., Kurz, A., Auerbach, S., Go, R., Sadovnick, D., Duara, R.,
Zimmerman, M., Mattia, J.I., 2001. A self-report scale to help make psychiatric
Kukull, W.A., Chui, H., Edeki, T., 2003. Depression as a risk factor for Alzheimer
diagnoses: the Psychiatric Diagnostic Screening Questionnaire. Arch. Gen.
disease: the MIRAGE Study. Arch. Neurol. 60 (5), 753–759.
Psychiatry. 58 (8), 787–794.