Archives of Gerontology and Geriatrics 78 (2018) 71–80

Contents lists available at ScienceDirect

Archives of Gerontology and Geriatrics

journal homepage: www.elsevier.com/locate/archger

Efficacy of group psychotherapy for geriatric depression: A systematic T

review
⁎
Tavares L.R.a, , Barbosa M.R.b
a
Faculty of Psychology and Education Sciences, University of Porto, Rua Alfredo Allen, 4200-135, Porto, Portugal
b
Center for Psychology at the University of Porto, Faculty of Psychology and Education Sciences, University of Porto, Rua Alfredo Allen, 4200-135, Porto, Portugal

A R T I C LE I N FO A B S T R A C T

Keywords: Background: Geriatric depression is a common and debilitating psychopathology, but evidence supports the
Geriatric depression efficacy of psychotherapy in its treatment. Group therapy provides advantages over individual interventions.
Group psychological therapy However, only three systematic reviews have focused specifically on the efficacy of group therapy for geriatric
Older adults depression.
Systematic review
Objective: To ascertain the effects of group psychotherapy on geriatric depression in people aged 60 years and
older, compared with alternative treatments or no treatment.
Data sources: A systematic review of English, Portuguese, and Spanish studies using the EBSCOhost Research and
Science Direct databases (2011–2017). Additional studies were identified through reference lists. Search terms
included group therapy, group psychotherapy, older adults, elderly, depressive disorder, geriatric depression,
and depression in the elderly.
Review methods: The researcher screened any study designs concerning the effects of any paradigm of group
therapy on geriatric depression versus alternative interventions or no treatment. Relevant data, including in-
dicators of risk of bias, were extracted.
Data synthesis: Nine studies were reviewed. Reminiscence therapy and cognitive-behavioral therapy are viable
group interventions for geriatric depression, and were significantly superior to most controls. Conclusions about
the long-term effects were unclear. Significant improvements were obtained for different intervention durations
and facilitators, and with participants of different nationalities and age. Most studies recruited participants from
the community, which limited generalizability. Group therapy also resulted in improvements in other psycho-
logical variables.
Conclusions: Group therapy can significantly improve geriatric depression. Improvements were found across a
variety of settings, protocols, participant characteristics, and for several psychological domains.

1. Introduction increased risk of morbidity and self-neglect, and deficits at physical,

cognitive, and social levels. This results in increased mortality rates
1.1. Rationale (Blazer, 2003; Casey, 2012; Fiske et al., 2009; Funnell, 2010). GD also
tends to be more chronic and persistent than depression in younger
1.1.1. Geriatric depression adults (Blazer, 2003), and often exists associated with medical condi-
The increase of human lifespan and the decrease in birthrates has tions typical of late adulthood (Blazer, 2003; Casey, 2012). Depressed
culminated in the aging of modern societies (Laidlaw, Thompson, Dick- elderly clients are a common presence in primary health care facilities,
Siskin, & Gallagher-Thompson, 2003). Portugal has one of the largest which results in increased costs and expenses. Thus, GD is a serious
aging rates in Europe (140%), and 18,7%–28,6% of the population is threat to older adults’ mental health, but also to society itself (Casey,
aged 65 years or older (Instituto Nacional de Estatística, 2016). 2012).
Late adulthood is associated with a high prevalence of psycho-
pathology, of which geriatric depression (GD) is the most common 1.1.2. Intervention
(Casey, 2012; Fiske, Wetherell, & Gatz, 2009; Laidlaw et al., 2003). With adequate treatment, a high percentage of depressed older
Depressed older adults have decreased functionality and well-being, adults can fully recover (Casey, 2012). These clients tend to have

⁎
Corresponding author.
E-mail addresses: mipsi12074@fpce.up.pt (L.R. Tavares), raquel@fpce.up.pt (M.R. Barbosa).

https://doi.org/10.1016/j.archger.2018.06.001
Received 14 March 2018; Received in revised form 7 May 2018; Accepted 9 June 2018
0167-4943/ © 2018 Elsevier B.V. All rights reserved.
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

positive attitudes towards psychotherapy and may prefer it instead of limited to group RT and to long-term care residents. The most recent
pharmacological treatment (Laidlaw, 2006). Several systematic reviews systematic review with a broad scope was the one conducted by Krishna
have demonstrated that psychotherapy is an effective evidence-based et al., 2011, new results and conclusions have become available.
practice to treat GD (e.g., Apóstolo, Bobrowicz-Campos, Rodrigues, Therefore, the aim of the present study was to expand on currently
Castro, & Cardoso, 2016; Jonsson et al., 2016), while others have de- available knowledge by systematically reviewing and synthesizing
monstrated the efficacy of group therapy in the treatment of depression published and unpublished materials. We reviewed any study designs,
in younger adults (e.g., Huntley, Araya, & Salisbury, 2012; Okumura & to ascertain the efficacy of group psychotherapy on depressive disorders
Ichikura, 2014). or depressive symptoms in people aged 60 years and older, compared
While group and individual modalities share characteristics, such as with alternative treatments or no treatment.
offering the client a safe haven where to expose their concerns and
proportioning a therapeutic alliance, group therapy shows some ad- 2. Methods
vantages. In a group, one can socialize with peers who share similar
symptoms, and there is opportunity to increase one’s altruism and 2.1. Eligibility criteria
empathy and to feel useful by helping others. As well, group therapy
results in a more cost-effective professionals/clients ratio (Agronin, 2.1.1. Population
2009). In particular, group therapy is an opportunity for depressed Participants must be 60 years or older, and formally diagnosed with
older adults, who often live isolated and have lost significant relation- a depressive disorder according to the definitions by the American
ships, capacities, or occupations, to once again feel connected with and Psychiatric Association or the World Health Organization, or have
useful to the society, as well as to (re)discover a meaning for their life significant depressive symptoms as measured with a validated scale. We
(Floyd & Scogin, 1998). excluded studies including participants with comorbid psychopatholo-
Group therapy rationales are similar for depressed young and older gies except anxiety disorders (due to common co-occurrence with de-
adults, though, for the latter, adaptations may be required such as pression; Kaufman & Charney, 2000), and studies including participants
shorter session times, written materials with bigger fonts, or repetition with any degree of cognitive impairment.
of instructions. However, older adults shouldn't immediately be taken
as clients with deficits or with reduced mental capacity, and the group 2.1.2. Interventions
protocols should be adapted only after consideration of the group's Any group intervention based on an explicit psychological rationale.
idiosyncrasies (Laidlaw et al., 2003; Laidlaw, 2006). We excluded studies where group therapy was administrated and as-
Some systematic reviews have investigated the efficacy of group sessed together with another treatment. There was one exception
therapy on GD, but not specifically. Engels and Vermey (1997); (Wuthrich, Rapee, Kangas, & Perini, 2016). We included this study
Bohlmeijer, Smit, and Cuijpers, (2003), and Franck, Molyneux, and because psychotropic medication was a constant to all participants,
Parkinson, (2016) included both individual and group modalities of they should refrain from making changes to their medication status
treatment. Pinquart and Sörensen (2001) also included individual and throughout the study, and it was monitored at post-test and follow-up
group treatments, and some included studies contemplated interven- assessments.
tions with non-depressed older adults. Payne and Marcus (2008) did not
consider GD as primary outcome. Although it wasn’t a main objective of 2.1.3. Comparator
their respective study, Cuijpers, van Straten, and Smit, (2006) and Any comparator.
Pinquart, Duberstein, and Lyness, (2007) concluded that individual and
group therapies are similarly effective in the treatment of GD. 2.1.4. Outcome and measures
To our knowledge, only three systematic reviews have focused The primary outcome considered was change in depressive symp-
specifically on the efficacy of group therapy on GD. Gorey and Cryns toms or remission. Psychological secondary outcomes assessed in the
(1991) analyzed 19 studies conducted between 1967 and 1988. The included studies were also examined. Any validated measure was ac-
main paradigms considered were cognitive-behavioral therapy (CBT) ceptable.
and psychodynamic therapy (PT). The mean global effect of group work
with depressed older adults was statistically and clinically significant, 2.1.5. Study design
and was homogeneous across all older age cohorts, group work dura- Any study design, except systematic reviews and meta-analyses.
tion, and clinical paradigms. The main limitation of this review was the
low methodological quality of the studies included. 2.1.6. Setting
Krishna et al. (2011) reviewed published and unpublished materials Any setting.
until 2009, and 6 studies met inclusion criteria for meta-analysis. All
were randomized controlled trials (RCT) of group CBT. Group CBT was 2.1.7. Language
significantly superior when compared to waitlist, with a modest overall Studies in Portuguese, English, or Spanish.
effect size (maintained at follow-up), and such benefits, when compared
to active controls, did not reach statistical significance. The main lim- 2.1.8. Publication type
itations of this review were the small number of eligible studies, the Studies published in peer-reviewed journals and unpublished ma-
small sample sizes, the relatively high attrition rate, and the hetero- terials.
geneity of the interventions.
Syed Elias, Neville, and Scott, (2015) investigated the efficacy of 2.2. Information sources
group reminiscence therapy (RT) on GD, among other primary out-
comes, by reviewing any design studies published between 2002 and The search was performed by the researcher between May and
2014. Concerning GD, 8 quasi-experimental trials were analyzed. Five August 2017. Studies dating between 2011 and July 2017 were iden-
of these found that group RT was effective in reducing GD, while three tified by searching the EBSCOhost Research and Science Direct data-
studies yielded non-significant findings. Only integrative reminiscence bases, and by scanning the reference lists of relevant systematic re-
therapy (IRT) was significantly effective in reducing GD. Limitations of views. In the EBSCOhost Research databases, we searched American
this review included small sample sizes, high attrition rates, insufficient Doctoral Dissertations, PsycINFO, Psychology and Behavioral Sciences
evidence about long-term effects, and use of passive controls only. Collection, PsycARTICLES, MEDLINE, MedicLatina, Fonte Acadêmica,
Although the review of Syed Elias et al. is the most recent, it was CINAHL Plus with Full Text, and Academic Search Complete.

72
 Table 1
Summary of studies about group therapy for geriatric depression.
Author/ Year/ Country Study design (follow-up Setting/ Sample Interventions (n) Nr. of sessions/ Facilitator(s) Outcome Main results
period) size (n)/ Mean age measures

Chueh and Chang, (2014)/ Quasi-experimental study Veterans’ RT: 11 4/ Mental health professionals GDS Group RT significantly
Taiwan (3 months + 6 months) nursing home/ 21 Waitlist: 10 decreased GD (p < .01) at post-test, and the effect was
L.R. Tavares, M.R. Barbosa

participants/ 82 years maintained at 3 months and 6 months follow-ups.

Group therapy was significantly more effective than control.
Drăghic, (2012)/ Romania Quasi-experimental study Unspecified/ 30 Experiential 8/ Unspecified HDRS No statistical analyses were reported.
(6 months) participants psychotherapy: 30 HARS
No control group. GHQ
SDS
WMS
Duyan, et al. (2017)/ Quasi-experimental study Nursing home/ 30 RT: 16 10/ Social work professionals MOSES Group RT did not significantly
Turkey participants No-treatment: 14 GDS-SF decrease GD (F3, 56 = 0.84, p = .48), but significantly
improved withdrawn behavior (F3,56 = 4.52, p < .05).
García-Peña et al. (2015)/ Randomized controlled Primary care unit/ 81 CBT: 41 12/ Nurses PHQ-9 Group CBT resulted in a decrease equal or superior to 5 points
Mexico trial participants/ 70.8 years TAU: 40 SF-36 on the PHQ-9 for 56% of the participants, and a decrease
SAST between 4 to 1 points for 19.5% of the participants
Hanaoka et al. (2011)/ Quasi-experimental study Community/ 22 RT: 22 8/ Mental health GDS-SF Group RT significantly
Japan participants/ 76.6 years No control group. professionals + volunteers TMIG-IC decreased GD (p = .04), but had no significant effect on
LSI-K subjective well-being.
Huang et al. (2015)/ Randomized controlled Community/ 57 CBT: 18 12/ Nurses GDS-SF Group CBT significantly
Taiwan trial (3 months + 6 participants/ 76.5 years PFE: 19 6-minute walk decreased GD (p < .01) at post-test, but the effect became non-
months) No-treatment: 20 test significant at 3 months and 6 months follow-ups.
SF-36 Group therapy was significantly more effective than no-
ISSB treatment and equivalent to group PFE.
Lobo et al. (2012)/ Brasil Quasi-experimental study Unspecified/ 6 CBT: 6 7/ Unspecified MMSE No statistical analyses were reported.

73
participants No control group. BDI
BAI
GDS-SF
MAC-Q
McLaughlin and Randomized controlled Community/ 37 CBT: 18 6/ Mental health professionals CIDI There was a significant main effect of time on GD (p < .05),
McFarland, (2011)/ trial (3 months) participants General care: 19 GDS but not of groups. Group therapy was equivalent to control.
Australia WPSI The same pattern was found for psychosocial functioning
(p < .01).
Moral et al. (2015)/ Quasi-experimental study Community/ 34 IRT: 17 8/ Mental health professionals GDS Group IRT significantly
Dominican Republic participants/ 73.9 Waitlist: 17 RSES decreased GD (F1,32 = 19.47, p < .001, η2 = 0.378).
LIS Group therapy was significantly more effective than control.
LSI-A The same pattern was found for self-esteem, life satisfaction,
RPWBS integrity, and psychological well-being
Wang et al. (2014)/ Quasi-experimental study Community/ 24 CBT: 12 8/ Mental health professionals HDRS Group CBT significantly
Taiwan participants General care: 12 WHOQOL- decreased GD (p < .01).
BREF Group therapy was significantly more effective than control.
The effect of group CBT on overall quality of life was not
significant when adjusted for depression.
Wu et al. (2012)/ Taiwan Quasi-experimental study Veterans’ Instrumental reminiscence 12/ Unspecified GDS-SF Group instrumental reminiscence significantly decreased GD
nursing home/ 37 psychotherapy: 17 LSI-A (p < .01) and increased life satisfaction and self-esteem
participants Unspecified: 20 RSES (p < .01 for both).
Wuthrich et al. (2016)/ Randomized controlled Community/ 133 CBT: 76 11/ Graduate students ADIS There was a significant time × group interaction for the mean
Australia trial (6 months) participants/ 67.4 years Active control: 53 in clinical psychology GDS severity of mood disorders (F2,241.570 = 3.412, p = .035), with
GAI group CBT being superior at post-test (t120 = −3.601,
WHOQOL- p < .001).
BREF There was a significant main effect of time on self-report
measures of GD, but no significant time x group interaction.
(continued on next page)
Archives of Gerontology and Geriatrics 78 (2018) 71–80
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

Anxiety Screening Test; TMIG-IC = TMIG Index of Competence; LSI-K = Life Satisfaction Index-K; ISSB = Inventory of Socially Supportive Behaviors Scales; MMSE = Mini Mental State Examination; BDI = Beck

RSES = Rosenberg Self-esteem Scale; LIS = Life Integration Scale; LSI-A = Life Satisfaction Index-A; RPWBS = Ryff Psychological Well-Being Scales; WHOQOL-BREF = World Health Organization Quality of Life
GDS = Geriatric Depression Scale; HDRS = Hamilton Depression Rating Scale; HARS = Hamilton Anxiety Rating Scale; GHQ = General Health Questionnaire; SDS = Social Dysfunction Subscale; WMS = Wechsler
Memory Scale; MOSES = Multidimensional Observation Scale for Elderly Subjects; GDS-SF = Geriatric Depression Scale Short Form; PHQ-9 = Patient Health Questionnaire; SF-36 = Short Form Health; SAST = Short

Depression Inventory; BAI = Beck Anxiety Inventory; MAC-Q = Memory Complaint Questionnaire; CIDI = Composite International Diagnostic Interview; WPSI = Washington Psychosocial Seizure Inventory;
2.3. Search strategy

Group therapy was significantly more effective than control

The same pattern was found for negative feelings, positive
The electronic search in the EBSCOhost Research databases used the
Boolean operators “group therapy or group psychotherapy” AND “older
Group RT significantly decreased GD (p < .001).

adults or elderly or geriatric” AND “depression or depressive disorder”.

In the Science Direct databases, we used the advanced search option for
the Boolean operators “geriatric depression or depression in the el-
derly” AND “group therapy or group psychotherapy”. We applied the
following filters, while selecting Journals Only: Medicine and dentistry,
Nursing and health professionals, Psychology, Social sciences.
feelings, and affect balance.

2.4. Study selection

The researcher screened the titles and abstracts for eligibility. All
Main results

materials of potential relevance, according to the inclusion criteria,

were retrieved in full text. The reference lists of relevant systematic
reviews and meta-analyses were screened for additional materials.

2.5. Data collection process

measures
Outcome

Data were extracted from each included study and inserted into a
RSES
GDS

ABS

table, created by the researcher according to Preferred Reporting Items

for Systematic reviews and Meta-Analyses (PRISMA; Liberati et al., 2009).

2.6. Data items

Measure-Brief; ADIS = Anxiety Disorders Interview Schedule for DSM-IV; GAI = Geriatric Anxiety Inventory; ABS = Affect Balance Sc.
Nr. of sessions/ Facilitator(s)

A summary of the information extracted and of the study char-

acteristics is presented in Table 1.

2.7. Risk of bias in individual studies

6/ Nurses

The risk of bias was assessed according to the standardized critical

appraisal instrument from the Joanna Briggs Institute Meta-Analysis of
Statistics Assessment and Review Instrument (JBI-MAStARI; The Joanna
Briggs Institute, 2014).

3. Results
Interventions (n)

General care: 66

3.1. Study selection

RT: 59

The search in the EBSCOhost Research and in the Science Direct

databases provided, respectively, 273 and 383 results, making a total of
656 results. After review of the titles and abstracts, while screening for
participants/ 69.4 years

duplicates, 476 results were discarded. The full text of 180 results was
size (n)/ Mean age

examined, including 5 identified from reference lists. After full text

Community/ 125
Setting/ Sample

assessment, while applying the selection criteria, 167 results were ex-
cluded, leaving a final total of 13 published studies (no unpublished
materials were considered eligible). Fig. 1 provides an overview of the
PRISMA strategy used to identify studies that met the selection criteria
(Liberati et al., 2009).
Quasi-experimental study
Study design (follow-up

3.2. Risk of bias within studies

Assessment of risk of bias is presented in Table 2. None of the se-

lected 13 studies fulfilled all the JBI-MAStARI criteria. We included
only the studies which fulfilled at least five out of ten criteria, excluding
period)

4 studies (Drăghici, 2012; Hanaoka, Muraki, Yamane, Shimizu, &

Okamura, 2011; Lobo et al., 2012; Wu, Chuo, & Wu, 2012). A total of 9
Zhou et al./ (2012)/ China

studies were analyzed.

Author/ Year/ Country
Table 1 (continued)

3.3. Study characteristics

Five studies had a quasi-experimental design (Chueh & Chang,

2014; Duyan, Şahin-Kara, Duyan, Özdemir, & Megahead, 2017; Moral,
Terrero, Galán, & Rodríguez, 2015; Wang, Tzeng, & Chung, 2014; Zhou
et al., 2012) and four were RCTs (García-Peña et al., 2015; Huang, Liu,

74
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

Fig. 1. Modified PRISMA flow diagram of study screening and selection.

Tsai, Chin, & Wong, 2015; McLaughlin & McFarland, 2011; Wuthrich discussion group. Huang et al. (2015) also included group physical
et al., 2016). Four studies included follow-up assessment (Chueh & fitness exercise (PFE). The number of sessions, frequency, and duration
Chang, 2014; Huang et al., 2015; McLaughlin & McFarland, 2011; of each session of these comparators’ protocols varied across studies.
Wuthrich et al., 2016), which varied between 3 and 6 months. Three GDS or GDS-SF were used in seven studies (Chueh & Chang, 2014;
studies were conducted in Taiwan (Chueh & Chang, 2014; Huang et al., Duyan et al., 2017; Huang et al., 2015; McLaughlin & McFarland, 2011;
2015; Wang et al., 2014), two in Australia (McLaughlin & McFarland, Moral et al., 2015; Wuthrich et al., 2016; Zhou et al., 2012). Other
2011; Wuthrich et al., 2016), one in Turkey (Duyan et al., 2017), one in scales used to assess GD included the HDRS (Wang et al., 2014) and the
Mexico (García-Peña et al., 2015), one in Dominican Republic (Moral PHQ-9 (García-Peña et al., 2015). As for interviews, McLaughlin and
et al., 2015), and one in China (Zhou et al., 2012). McFarland (2011) utilized the CIDI, and Wuthrich et al. (2016) utilized
Participants of all studies were aged 60 years or older and were the ADIS.
diagnosed with GD or had significant depressive symptoms as measured Psychological secondary outcomes included quality of life (García-
with a validated scale. Six studies included community dwellers (Huang Peña et al., 2015; Huang et al., 2015; Wang et al., 2014; Wuthrich et al.,
et al., 2015; McLaughlin & McFarland, 2011; Moral et al., 2015; Wang 2016), anxiety (García-Peña et al., 2015; Wuthrich et al., 2016), psy-
et al., 2014; Wuthrich et al., 2016; Zhou et al., 2012), one included chosocial functioning level and well-being (Duyan et al., 2017; Moral
participants from a primary care unit (García-Peña et al., 2015), one et al., 2015), self-esteem (Moral et al., 2015; Zhou et al., 2012), social
included nursing home residents (Duyan et al., 2017), and one included support (Huang et al., 2015), integrity (Moral et al., 2015), life sa-
a veteran’s nursing home residents (Chueh & Chang, 2014). tisfaction (Moral et al., 2015), and affect balance (Zhou et al., 2012).
The group intervention rationale was CBT in five studies (García-
Peña et al., 2015; Huang et al., 2015; McLaughlin & McFarland, 2011;
Wang et al., 2014; Wuthrich et al., 2016), RT in three studies (Chueh & 3.4. Results of individual studies
Chang, 2014; Duyan et al., 2017; Zhou et al., 2012), and IRT in one
study (Moral et al., 2015). The duration of the interventions varied Chueh and Chang (2014) explored the effects of group RT, com-
between 4 and 12 weeks. The group sessions were led by mental health pared to waitlist, on institutionalized male veterans. There was a sig-
professionals in four studies (Chueh & Chang, 2014; McLaughlin & nificant decrease in GDS scores in the intervention group but not in the
McFarland, 2011; Moral et al., 2015; Wang et al., 2014), by nurses in control group, maintained at 6 months follow-up. At pre-test, 81.81% of
three studies (García-Peña et al., 2015; Huang et al., 2015; Zhou et al., participants in the intervention group had a diagnosis of GD; at post-
2012), by social work professionals in one study (Duyan et al., 2017), test, this rate dropped to 9.09% (42.86% at follow-up). In the control
and by graduate students in clinical psychology in one study (Wuthrich group, the rate of GD was maintained around 80% (83.33% at follow-
et al., 2016). up).
Comparators included general care in three studies (McLaughlin & Duyan et al. (2017) explored the effects of group RT, compared to
McFarland, 2011; Wang et al., 2014; Zhou et al., 2012), no-treatment in no-treatment, on participants from a nursing home. There were no
two studies (Duyan et al., 2017; Huang et al., 2015), waitlist in two statistically significant differences in GDS-SF scores in the intervention
studies (Chueh & Chang, 2014; Moral et al., 2015), and treatment as group. However, behavioral changes were observed. These participants
usual (TAU) in one study (García-Peña et al., 2015). Wuthrich et al. also showed a significant reduction in withdrawn behavior. The parti-
(2016) had an active control, which consisted of a non-directive cipants in the control group had higher GDS-SF scores at post-test as-
sessment than they did at pre-test.

75
 L.R. Tavares, M.R. Barbosa

Table 2
Assessment of methodological quality of group therapy studies using JBI MAtSARI.
Criteria Chueh and Drăghici Duyan et al. García-Peña Hanaoka et al. Huang et al. Lobo et al. McLaughlin and Moral et al. Wang et al. Wu et al. Wuthrich et al. Zhou et al.
Chang (2012) (2017) et al. (2015) (2011) (2015) (2012) McFarland (2011) (2015) (2014) (2012) (2016) (2012)
(2014)

1. Was the assignment to treatment Y NA U Y NA Y NA Y N N U Y Y

groups truly random?
2. Were participants blinded to U NA U N NA U NA U U N U U U
treatment allocation?
3. Was allocation to treatment U NA U Y NA U NA U U N U Y U
groups concealed from the
allocator?
4. Were the outcomes of people Y N NA Y N NA NA U NA Y U Y U
who withdrew described and
included in the analysis?

76
5. Were those assessing outcomes U U U Y U Y U U U U U Y U
blind to the treatment
allocation?
6. Were the control and treatment Y NA Y Y NA Y NA Y Y Y U Y Y
groups comparable at entry?
7. Were groups treated identically Y NA Y Y NA Y NA Y Y Y U Y Y
other than for the named
interventions?
8. Were outcomes measured in the Y NA Y Y NA Y NA Y Y Y Y Y Y
same way for all groups?
9. Were outcomes measured in a Y Y Y Y Y Y Y Y Y Y Y Y Y
reliable way?
10. Was appropriate statistical Y NA Y Y Y Y U Y Y Y Y Y Y
analysis used?

Y = Yes, N = No, U = unclear, NA = not applicable.

Archives of Gerontology and Geriatrics 78 (2018) 71–80
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

García-Peña et al. (2015) explored the effects of group CBT, com- significant change in the control group. Self-esteem scores showed no
pared to TAU, on participants from a primary care unit. The main significant difference after the interventions.
outcome was the clinically relevant modification in the PHQ-9 scores,
and intervention effectiveness was defined as a decrease of at least 5 3.5. Syntheses of results
points in these scores. This was reported for 56.1% of the participants in
the intervention group, and for 30% in the control group. A decrease Group therapy resulted in significant improvements in GD symp-
between 4 to 1 points was reported for 19.5% of the participants in the toms in eight out of nine studies, the exception being Duyan et al.
intervention group, and for 22.5% in the control group. No change or (2017). Two studies reported no significant superiority of the group
increase of up to 5 points were reported for 24.3% of the participants in intervention compared to control, respectively: group CBT versus PFE,
the intervention group, and for 47.5% in the control group. The SF-36 and group CBT versus relaxation techniques. One study also reported no
and the SAST were also used, but no data was provided. superiority of group CBT versus a non-directive discussion group in self-
Huang et al. (2015) compared the effects of group CBT, group PFE, reported GD symptoms. Six studies reported that group therapy resulted
and no-treatment, on community dwellers. The GDS-SF scores sig- in improvements on GD which were significantly superior to those
nificantly decreased in the CBT and PFE groups, and, at post-test, there obtained with control intervention, respectively: group RT versus gen-
were more participants without a diagnosis of GD in both intervention eral care, group RT versus waitlist, group IRT versus waitlist, group CBT
groups than in the control group. The difference in the decrease of GDS- versus TAU, and group CBT versus general care. One study also reported
SF scores between both intervention groups was non-significant. Unlike superiority of group CBT versus a non-directive discussion group in GD
the group PFE effect, the group CBT effect became non-significant at 3- symptoms assessed by a professional.
month and 6-month follow-ups. The authors also discussed results ob- Significant effects on GD were obtained for group therapy protocols
tained for group CBT and group PFE for 6-minute walk distance, quality of different durations (ranging from 4 to 12 sessions) and led by dif-
of life, and social support. ferent facilitators (mental health professionals, nurses, and graduate
McLaughlin and McFarland (2011) had the main objective of ex- students in clinical psychology). The study with no significant im-
amining the effects of group CBT, compared with relaxation techniques, provements (Duyan et al., 2017) had 10 sessions, led by social work
on the seizure frequency of community dwellers with epilepsy. Since professionals.
the CIDI and the GDS were used to assess depression as well as dys- Significant effects on GD were obtained with participants from
thymia, we considered both as measures of GD. The effect of group CBT different settings (community dwellers, primary care unit, and veteran’s
on GD did not statistically differ from the effect of the control group nursing home) and of different nationalities (Taiwan, Australia, Mexico,
intervention. The same pattern of results was reported for psychosocial Dominican Republic, and China). Based on Funnell (2010)’s termi-
functioning. A significant Time x Group interaction on seizure fre- nology, significant effects were reported for “young-old” and for “old-
quency was also reported. old” participants. The study with no significant improvements recruited
Moral et al. (2015) explored the effects of group IRT, compared to participants from a nursing home in Turkey and did not provide their
waitlist, on community dwellers. There were no significant differences mean age.
in GDS scores in the control group, but a significant decrease was re- All studies where group therapy resulted in significant improve-
ported in the IRT group. The scores of self-esteem, life satisfaction, ments assessed GD with self-report instruments, except for one, using an
integrity, and psychological well-being significantly increased in the instrument rated by a professional. Two studies also utilized interviews.
IRT group, unlike in the control group. The Time x Group interactions Group therapy resulted, furthermore, in significant improvements
for all variables were significant. for anxiety, psychosocial functioning level and well-being, social sup-
Wang et al. (2014) explored the effects of group CBT, compared to port, integrity, life satisfaction, affect balance, and withdrawn beha-
general care, on community dwellers. Group CBT resulted in a sig- vior. Quality of life revealed mixed results: one study reported sig-
nificant decrease in HDRS scores. A significant difference from pre-test nificant general improvements, one study reported significant
to post-test assessment was reported for the psychological and social improvements only in the psychological and social domains, and one
domains of quality of life, but not for the physical and environmental study reported no significant improvements. Mixed results were also
domains. In the control group, GD did not improve and overall quality found for self-esteem: one study found significant improvements, and
of life decreased. one study did not.
Wuthrich et al. (2016) explored the effects of group CBT, compared
to an active control, on community dwellers. Considering the diagnosis 4. Discussion
of a primary (either depressive or anxiety) disorder, there was a sig-
nificant effect of time and a significant Time × Group interaction, 4.1. Summary of evidence
which indicated a significant improvement for group CBT participants
compared to the non-directive discussion group. This superiority be- 4.1.1. Quality of the studies
came non-significant at the 6 months follow-up. The same pattern of Of the nine studies analyzed, four were RCTs, which is considered
results was reported for the mean severity of all disorders. For the mean the best source of experimental evidence (The Joanna Briggs Institute,
severity of mood disorders only, there was a significant effect of time 2014). Cuijpers (2016) calculated that, to detect a medium effect size
and a significant Time × Group interaction, with group CBT being su- (Cohen’s d = .6), at least 90 trial participants are required, and that, to
perior at post-test assessment. The same pattern of results was reported detect a minimally relevant effect size (Cohen’s d = .24), this number is
for the mean severity of anxiety disorders only. A significant main effect at least 548 trial participants. In our review, only two studies involved a
of time on self-report measures of GD, anxiety, and overall quality of sample size with more than 100 participants. Therefore, more than the
life was also reported, but there were no significant Time x Group in- small number of RCTs, what warrants caution about our results is the
teractions on any of these scores. overall lack of power of the included studies.
Finally, Zhou et al. (2012) explored the effects of group RT, com- None of the studies fulfilled all the JBI-MAStARI criteria and, in all
pared to general care, on community dwellers. GDS scores of the par- of them, participant blinding to treatment allocation was unclear or not
ticipants in the RT group significantly decreased. In the control group, addressed. However, given the nature of psychotherapeutic interven-
GDS scores also decreased, but this result was significantly lower than tions, we acknowledged the practical difficulties of this procedure
that of the RT group. Negative feelings scores also decreased sig- (Barkham, Moller, & Pybis, 2017). Some studies which might have
nificantly more in the RT group, while the positive feelings and affect provided viable results were excluded due to risk of bias. Nonetheless,
balance scores significantly increased in the RT group and had no this improved the confidence with which conclusions can be drawn

77
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

from our review. effect was still superior at follow-up, or it was still significant but no
longer superior to controls, or it was no longer significant. The longest
4.1.2. Efficacy of group therapy for geriatric depression follow-up period considered was 6 months, and some studies reported
We expanded the knowledge established by previous reviews by drop-out of participants by that point. It's possible that lack of assess-
analyzing the most recent literature, by considering different compo- ment of attrition rates and reduced statistical power can explain the
nents of the studies, and by including different study designs and non-significant results. Another possibility, as suggested by Huang et al.
paradigms of group therapy. Of the nine studies analyzed, eight re- (2015), is that, often, participants aren't explicitly instructed nor taught
ported statistically significant improvements in GD after group therapy. how to maintain the positive changes after the group intervention is
Therefore, this modality of psychological intervention emerged as a finished. Group therapy protocols should start addressing this issue,
viable therapeutic option for older adults suffering from depressive while future research should continue to invest in follow-up assess-
disorders or depressive symptoms. In accordance to previous reviews, ments, include longer follow-up periods, and address the issue of at-
there was support for the efficacy of group RT and group CBT. trition rates.
Syed Elias et al. (2015) reported mixed results for group RT, which Given the variety identified in the duration of the interventions, we
is in accordance with our results. The authors also reported that IRT concluded that group therapy for GD is a malleable modality. Pinquart
was the only specific type of RT significantly effective in improving GD, et al. (2007) proposed that 7–12 sessions might provide optimal effi-
which was supported by our study through the work of Moral et al. cacy, and we identified positive results with interventions ranging from
(2015). This may be because IRT promotes the review of positive and 4 to 12 sessions. This suggests that the protocols can be adapted to both
negative past events, to meaningfully connect past and current mem- the professionals’ and the clients’ needs, without such modifications
ories (Hallford, Mellor, & Cummins, 2012, cit. in Syed Elias et al., 2015). affecting the intervention’s efficacy. The same applies to the group
The cognitive restructuring inherent to IRT may explain why it was sessions’ facilitators, since we found significant improvements after
more effective than other types of group RT. group therapy led by professionals from different areas.
Group CBT, as the individual modality, proposes that the depressive The extent to which our results can be generalized is somewhat
humor derives from maladaptive thought patterns (Beck, Rush, Shaw, & limited, because most of the included studies recruited participants
Emery, 1979). Thus, its primary goal is to identify and change dys- from the community. These authors alerted for the healthier and
functional beliefs, thoughts, and behaviors (Casey, 2012; Floyd & wealthier condition of these persons, compared to the general elderly
Scogin, 1998), while considering the meanings that clients attribute to population. However, the study of Chueh and Chang (2014) is worthy
their life experiences (Beck et al., 1979; Laidlaw, 2006). Typically, of attention. Their participants were recruited from a war veterans’
group CBT protocols include three modules: cognitions, behaviors, and nursing home, with idiosyncratic life experiences and affected by sev-
social interactions, and their respective impact on depressive humor. eral stress factors, which can culminate in severe depressive symptoms.
Gorey and Cryns (1991) concluded that group CBT and group PT Nevertheless, group therapy improved GD in this study, with effects
were statistically and clinically effective interventions for GD and that significantly superior to waitlist at post-test and 6 months follow up.
improvements were homogeneous across all older age cohorts studied Group therapy also yielded significant improvements for older
and group work duration. Our results obtained for group CBT were in adults of different nationalities. While the aging of societies is a
accordance, although no group PT studies were eligible. Krishna et al. worldwide phenomenon, a majority of the published literature consists
(2011) also concluded that group CBT was an effective intervention for of samples from the United States or Europe (Neto, 2002), and minority
GD. However, the effect size obtained was non-significant when com- groups in such countries tend to be underserved when reaching out to
pared to active controls. Floyd and Scogin (1998) discussed the mental health care facilities (Barkham et al., 2017). Our review showed
“common factors” shared by different paradigms of group therapy, that different protocols and paradigms of group therapy can be used for
which may be the reason for the positive results obtained with older GD with good results in culturally diverse clients, although it was not
adults, as opposed to specific factors from any paradigm. In our review, possible to identify any features specific to any given country or culture.
the study of Wuthrich et al. (2016) compared the efficacy of group CBT Future research should focus on trying to identify such potential spe-
with a non-directive discussion group which served as control for cificities, on promoting group therapy protocols which are mindful of
generic psychotherapy skills and typical group processes. Both inter- culturally diverse clients, and on examining their efficacy.
ventions resulted in significant improvements in self-reported GD Group therapy helped improve GD even for the “old-old” partici-
symptoms, among other outcomes, and this evidenced the relevance of pants. Due to the increase in human lifespan, societies started ac-
general group therapeutic processes. However, group CBT resulted in knowledging a fourth age, starting after 75–80 years of age. Whilst it is
significantly greater improvements in GD symptoms assessed by a possible to remain physically and mentally healthy at such prolonged
professional, among other outcomes, demonstrating some benefits be- age, it is also undeniable that functional incapacities, chronic diseases,
yond general, non-specific aspects of group therapy. and overall stress factors may result in impoverished quality of life and
Our review demonstrated that group therapy resulted in improve- higher risk for psychopathology (Baltes & Smith, 2003). Our results
ments in GD which were significantly superior to those obtained with demonstrated that, despite the inevitable consequences at several levels
no-treatment, waitlist, general care, and TAU. Huang et al. (2015) as- of reaching such an old age, group therapy can still be of use for these
certained that group CBT and PFE were equivalent in ameliorating GD persons.
at post-test assessment, and that the effects of group CBT became non- Group therapy significantly improved GD symptoms as perceived by
significant at follow-up. The authors suggested explanations such as the participants (self-report scales) and as perceived by professionals
both interventions having a group format, participants not maintaining (scales rated by the professional and interviews). While there is in-
the CBT practice, and the subsequent decrease in social interaction after creased risk of bias when the same professional assesses outcomes and
the CBT group was ended. McLaughlin and McFarland (2011) also re- conducts the intervention (The Joanna Briggs Institute, 2014), depres-
ported that group CBT and relaxation techniques were equivalent in sive symptoms are also associated with negative attitudes and ex-
ameliorating GD. Although the authors did not explore this possibility, pectations about the self and the future (Beck et al., 1979). Group
we alert that relaxation techniques are often a component of CBT therapy providing significant improvements in GD, when the symptoms
protocols. Therefore, this control group limited the conclusions ob- were assessed by different methods, is further evidence of the utility of
tained. this modality of treatment.
As for the long-term effects of group therapy on GD, our conclusions Finally, group therapy also resulted in improvements in several
were unclear. Most included studies did not conduct a follow-up as- variables associated with mental health and well-being, which adds to
sessment, and those which did yielded mixed results, i.e., the significant its value. The discrepancy observed in quality of life may be explained

78
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

by the fact that the authors assessed it with different measures. Both Baltes, P. B., & Smith, J. (2003). New frontiers in the future of aging: From successful
studies providing contradictory results in self-esteem assessed it with aging of the young old to the dilemmas of the fourth age. Gerontology, 49, 123–135.
http://dx.doi.org/10.1159/000067946.
the RSES. Both also recruited participants from the community, and Barkham, M., Moller, N. P., & Pybis, J. (2017). How should we evaluate research on
followed the RT paradigm. However, Moral et al. (2015) utilized group counselling and the treatment of depression? A case study on how the National
IRT, which was already discussed as the type of RT providing superior Institute for Health and Care Excellence’s draft 2018 guideline for depression con-
sidered what counts as best evidence. Counselling and Psychotherapy Research, 1–16.
effect. http://dx.doi.org/10.1002/capr.12141.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of depression.
4.2. Limitations New York: Guilford.
Blazer, D. G. (2003). Depression in late life: Review and commentary. Journal of
Gerontology, 58A(3), 249–265.
Our systematic review has several limitations. This study was a Bohlmeijer, E., Smit, F., & Cuijpers, P. (2003). Effects of reminiscence and life review on
dissertation to obtain the Master’s degree in Clinical and Health late-life depression: A meta-analysis. International Journal of Geriatric Psychiatry, 18,
1088–1094. http://dx.doi.org/10.1002/gps.1018.
Psychology, which forced it to be conducted by only one researcher and
Casey, D. A. (2012). Depression in the elderly: A review and update. Asia-Pacific
imposed time limitations. Psychiatry, 4(3), 160–167. http://dx.doi.org/10.1111/j.1758-5872.2012.00191.x.
A relatively small number of studies was included, none of which Chueh, K., & Chang, T. (2014). Effectiveness of group reminiscence therapy for depressive
had ideal methodological quality and most of which lacked adequate symptoms in male veterans: 6-month follow-up. International Journal of Geriatric
Psychiatry, 29, 377–383. http://dx.doi.org/10.1002/gps.4013.
statistical power. Only published studies, written in English, Cuijpers, P. (2016). Are all psychotherapies equally effective in the treatment of adult
Portuguese, or Spanish, were ultimately selected. We excluded studies depression? The lack of statistical power of comparative outcome studies. Evidence
recruiting participants with comorbid psychopathologies (except an- Based Mental Health, 19, 39–42. http://dx.doi.org/10.1136/eb-2016-102341.
Cuijpers, P., van Straten, A., & Smit, F. (2006). Psychological treatment of late-life de-
xiety disorders) and with any degree of cognitive impairment. pression: A meta-analysis of randomized controlled trials. International Journal of
However, it is rare that a person suffers from a well-defined disorder Geriatric Psychiatry, 21, 1139–1149. http://dx.doi.org/10.1002/gps.1620.
without overlapping symptoms (Barkham et al., 2017). Besides, symp- Drăghici, R. (2012). Experiential psychotherapy in geriatric groups. Social and Behavioral
Sciences, 33, 979–983. http://dx.doi.org/10.1016/j.sbspro.2012.01.268.
toms of GD are often associated with, or even mistaken for, those of Duyan, V., Şahin-Kara, G., Duyan, G. C., Özdemir, B., & Megahead, H. A. (2017). The
dementia (Kolb & Wishaw, 2015). effects of group work with institutionalized elderly persons. Research on Social Work
We only assessed the efficacy of group CBT and group RT, lacking Practice, 27(3), 366–374. http://dx.doi.org/10.1177/1049731516654572.
Engels, G. I., & Vermey, M. (1997). Efficacy of nonmedical treatments of depression in
data about other paradigms. Finally, our review included only quanti- elders: A quantitative analysis. Journal of Clinical Geropsychology, 3, 17–35.
tative studies. Qualitative and mixed designs could be useful to un- Fiske, A., Wetherell, J. L., & Gatz, M. (2009). Depression in older adults. Annual Review of
derstand what interventions work for depressed older adults in actual Clinical Psychology, 5, 363–389. http://dx.doi.org/10.1146/annurev.clinpsy.032408.
153621.
practice, why they work (or not), and the attitudes of these clients to-
Floyd, M., & Scogin, F. (1998). Cognitive-behavior therapy for older adults: How does it
wards said interventions (Barkham et al., 2017). work? Psychotherapy, 35(4), 459–463.
Franck, L., Molyneux, N., & Parkinson, L. (2016). Systematic review of interventions
4.3. Conclusions addressing social isolation and depression in aged care clients. Quality of Life
Research, 25(6), 1395–1407. http://dx.doi.org/10.1007/s11136-015-1197-y.
Funnell, E. (2010). Depression in the elderly. InnovAiT, 3(4), 199–208. http://dx.doi.org/
Group therapy is a viable option to improve the symptoms of GD 10.1093/innovait/inp231.
and to ameliorate the mental health and well-being of older adults. García-Peña, C., Vázquez-Estupiñan, F., Avalos-Pérez, F., Jiménez, L., Sánchez-Garcia, S.,
& Juárez-Cedillo, T. (2015). Clinical effectiveness of group cognitivebehavioural
Group CBT constitutes a valid and reliable paradigm, while group RT therapy for depressed older people in primary care: A randomised controlled trial.
seems to also be recommendable despite less consistent results. Group Salud Mental, 38(1), 33–39.
therapy provides improvements across a variety of settings, protocols, Gorey, K. M., & Cryns, A. G. (1991). Group work as interventive modality with the older
depressed client: A meta-analytic review. Journal of Gerontological Social Work, 16(1-
and participant characteristics, and benefits several psychological do- 2), 137–157. http://dx.doi.org/10.1300/J083v16n01_11.
mains. Given the variable risk of bias present in the reviewed studies, Hanaoka, H., Muraki, T., Yamane, S., Shimizu, H., & Okamura, H. (2011). Testing the
our conclusions warrant some caution and should not be taken as ab- feasibility of using odors in reminiscence therapy in Japan. Physical & Occupational
Therapy in Geriatrics, 29(4), 287–299. http://dx.doi.org/10.3109/02703181.2011.
solute. Nevertheless, even when statistically significant improvements
628064.
are not obtained, the clinical benefits may justify the use of group Huang, T., Liu, C., Tsai, Y., Chin, Y., & Wong, C. (2015). Physical fitness exercise versus
therapy, because any small improvements that help decrease the suf- cognitive behavior therapy on reducing the depressive symptoms among community-
dwelling elderly adults: A randomized controlled trial. International Journal of Nursing
fering of depressed elderly (or even that halt further worsening of the
Studies, 52, 1542–1552. http://dx.doi.org/10.1016/j.ijnurstu.2015.05.013.
symptoms) are relevant and desirable. Besides to further validate our Huntley, A. L., Araya, R., & Salisbury, C. (2012). Group psychological therapies for de-
conclusions, to clarify aspects not yet addressed (e.g., comparison with pression in the community: Systematic review and meta-analysis. The British Journal
other treatment modalities), and to address our limitations, future re- of Psychiatry, 200, 184–190. http://dx.doi.org/10.1192/bjp.bp.111.092049.
Instituto Nacional de Estatística (2016). Portal do Instituto Nacional de Estatística.
search should contemplate understudied factors such as researcher al- Proporção da população residente com 65 ou mais anos de idade (%) por Local de
legiance or therapist effects (Barkham et al., 2017). residência (NUTS - 2013) e Tipologia de áreas urbanas; Anual (4 de fevereiro de 2016)
Retrieved fromhttp://www.ine.pt/.
Jonsson, U., Bertilsson, G., Allard, P., Gyllensvärd, H., Söderlund, A., Tham, A., et al.
Funding sources (2016). Psychological treatment of depression in people aged 65 years and over: A
systematic review of efficacy, safety, and cost-effectiveness. PLoS ONE, 11(8), http://
This research did not receive any specific grant from funding dx.doi.org/10.1371/journal.pone.0160859.
Kaufman, J., & Charney, D. (2000). Comorbidity of mood and anxiety disorders.
agencies in the public, commercial, or not-for-profit sectors. Depression and Anxiety, 12(1), 69–74.
Kolb, B., & Wishaw, I. Q. (2015). Fundamentals of human neuropsychology (7th ed.). New
Declarations of interest York: Worth Publications.
Krishna, M., Jauhari, A., Lepping, P., Turner, J., Crossley, D., & Krishnamoorthy, A.
(2011). Is Group psychotherapy effective in older adults with depression? A sys-
None. tematic review. International Journal of Geriatric Psychiatry, 26, 331–340. http://dx.
doi.org/10.1002/gps.2546.
Laidlaw, K. (2006). Cognitive behavior therapy with older adults. In S. H. Qualls, & B. G.
References
Knight (Eds.). Psychotherapy for depression in older adults.. New Jersey: John Wiley &
Sons, Inc.
Agronin, M. (2009). Group therapy in older adults. Current Psychiatry Reports, 11(1), Laidlaw, K., Thompson, L. W., Dick-Siskin, L., & Gallagher-Thompson, D. (2003).
27–32. Cognitive behaviour therapy with older people. West Sussex, England: John Wiley &
Apóstolo, J., Bobrowicz-Campos, E., Rodrigues, M., Castro, I., & Cardoso, D. (2016). The Sons, Ltd.
effectiveness of non-pharmacological interventions in older adults with depressive Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gøtzsche, P. C., Ioannidis, J., et al.
disorders: A systematic review. International Journal of Nursing Studies, 58, 59–70. (2009). The PRISMA statement for reporting systematic reviews and meta-analyses of
http://dx.doi.org/10.1016/j.ijnurstu.2016.02.006. studies that evaluate health care interventions: Explanation and elaboration. Journal

79
L.R. Tavares, M.R. Barbosa Archives of Gerontology and Geriatrics 78 (2018) 71–80

of Clinical Epidemiology, 62, e1–e34. http://dx.doi.org/10.1016/j.jclinepi.2009.06. Pinquart, M., Duberstein, P. R., & Lyness, J. M. (2007). Effects of psychotherapy and other
006. behavioral interventions on clinically depressed older adults: A meta-analysis. Aging
Lobo, B., Rigoli, M., Sbardelloto, G., Rinaldi, J., Argimon, I., & Kristensen, C. (2012). & Mental Health, 11(6), 645–657. http://dx.doi.org/10.1080/13607860701529635.
Terapia cognitivo‑comportamental em grupo para idosos com sintomas de ansiedade Syed Elias, S. M., Neville, C., & Scott, T. (2015). The effectiveness of group reminiscence
e depressão: Resultados preliminares. Psicologia: teoria e prática, 14(2), 116–125. therapy for loneliness, anxiety and depression in older adults in long-term care: A
McLaughlin, D. P., & McFarland, K. (2011). A randomized trial of a group based cognitive systematic review. Geriatric Nursing, 36(5), 372–380. http://dx.doi.org/10.1016/j.
behavior therapy program for older adults with epilepsy: The impact on seizure gerinurse.2015.05.004.
frequency, depression and psychosocial well-being. Journal of Behavioral Medicine, The Joanna Briggs Institute (2014). Joanna briggs institute reviewers’ manual: 2014 edition.
34, 201–207. http://dx.doi.org/10.1007/s10865-010-9299-z. The University of Adelaide, South Australia: The Joanna Briggs Institute.
Moral, J., Terrero, F., Galán, A., & Rodríguez, T. (2015). Effect of integrative reminiscence Wang, C., Tzeng, D., & Chung, W. (2014). The effect of early group psychotherapy on
therapy on depression, well-being, integrity, self-esteem, and life satisfaction in older depressive symptoms and quality of life among residents of an apartment building for
adults. The Journal of Positive Psychology, 10(3), 240–247. http://dx.doi.org/10. seniors. Psychogeriatrics, 14, 38–46. http://dx.doi.org/10.1111/psyg.12037.
1080/17439760.2014.936968. Wu, L. F., Chuo, L. J., & Wu, S. T. (2012). The effect of group instrumental reminiscence
Neto, F. (2002). Psicologia intercultural. Lisboa: Universidade Aberta. therapy in older single veterans who live in a veterans home in Taiwan. International
Okumura, Y., & Ichikura, K. (2014). Efficacy and acceptability of group cognitive beha- Journal of Geriatric Psychiatry, 27, 107–108 10.002/gps.2665.
vioral therapy for depression: A systematic review and meta-analysis. Journal of Wuthrich, V. M., Rapee, R. M., Kangas, M., & Perini, S. (2016). Randomized controlled
Affective Disorders, 164, 155–164. http://dx.doi.org/10.1016/j.jad.2014.04.023. trial of group cognitive behavioral therapy compared to a discussion group for co-
Payne, K. T., & Marcus, D. K. (2008). The efficacy of group psychotherapy for older adult morbid anxiety and depression in older adults. Psychological Medicine, 46, 785–795.
clients: A meta-analysis. Group Dynamics: Theory, Research, and Practice, 12(4), http://dx.doi.org/10.1017/S0033291715002251.
268–278. http://dx.doi.org/10.1037/a0013519. Zhou, W., He, G., Gao, J., Yuan, Q., Feng, H., & Zhang, C. K. (2012). The effects of group
Pinquart, M., & Sörensen, S. (2001). How effective are psychotherapeutic and other reminiscence therapy on depression, self-esteem, and affect balance of Chinese
psychoscial interventions with older adults? A meta-analysis. Journal of Mental Health community-dwelling elderly. Archives of Gerontology and Geriatrics, 54, e440–e447.
and Aging, 7(2), 207–242. http://dx.doi.org/10.1016/j.archger.2011.12.003.

80