Non-pharmacological Approaches to Depressed Elderly With No or Mild Cognitive Impairment in Long-Term Care Facilities. A Systematic Review of the Literature

Introduction: Compared to old people who live at home, depressive symptoms are more prevalent in those who live in long-term care facilities (LTCFs). Different kinds of non-pharmacological treatment approaches in LTCFs have been studied, including behavioral and cognitive-behavioral therapy, cognitive bibliotherapy, problem-solving therapy, brief psychodynamic therapy and life review/reminiscence. The aim of the current review was to systematically review non-pharmacological treatments used to treat depressed older adults with no or mild cognitive impairment (as described by a Mini Mental State Examination score > 20) living in LTCFs. Methods: A research was performed on PubMed and Scopus databases. Following the Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) flowchart, studies selection was made. The quality of each Randomized Controlled Trial was scored using the Jadad scale, Quasi-Experimental Design studies and Non-Experimental studies were scored based on the Newcastle-Ottawa Scale (NOS) Results: The review included 56 full text articles; according to the type of intervention, studies were grouped in the following areas: horticulture/gardening (n = 3), pet therapy (n = 4), physical exercise (n = 9), psychoeducation/rehabilitation (n = 15), psychotherapy (n = 3), reminiscence and story sharing (n = 14), miscellaneous (n = 8). Discussion and Conclusion: Despite mixed or negative findings in some cases, most studies included in this systematic review reported that the non-pharmacological interventions assessed were effective in the management of depressed elderly in the LTCFs context. Regrettably, the limitations and heterogeneity of the studies described above hinder the possibility to generalize and replicate results.


INTRODUCTION
Different kinds of non-pharmacological treatment approaches in LTCFs have been applied, with results suggesting a high variability on the subpopulations of elderly involved, settings, treatment models and outcomes (8)(9)(10). Several psychotherapeutic approaches have been studied and described in previous reviews, including behavioral and cognitive-behavioral therapy, cognitive bibliotherapy, problemsolving therapy, brief psychodynamic therapy, and life review/reminiscence (11)(12)(13)(14), Reminiscence Therapy (RT) (15), light therapy (16), music-art-creative therapy (17), and pet therapy (18). Nonetheless, only few of these reviews specifically focused on the specific population of non-cognitively impaired elderly living in nursing homes.
The aim of the current review was to systematically review non-pharmacological treatments used to treat depressed older adults with no or mild cognitive impairment (as described by a Mini Mental State Examination [MMSE] score > 20) living in LTCFs, in order to offer an up-to-date synthesis of the available literature about this topic.

METHODS
Our research question was the following: Are nonpharmacological approaches effective in addressing depression in elderly people with no or mild cognitive impairment residing in LTCFs?
Our PICO (20) was: Population Elderly people (age over 65 years old) residing in LTCFs, with no or mild cognitive impairment

Intervention
Any non-pharmacological approach  a third reviewer (C.G.) resolved any discrepancies that emerged between the reviewers.
The following inclusion/exclusion criteria were established for the selection of studies: Patients' age over 65 years old; patients living in nursing home residence or similar structures (LTCFs); diagnosis of depression, either based on clinical assessment or on the administration of questionnaires or scales; no or mild cognitive impairment (as described by a MMSE score > 20, or by a score on the Mini-Cog ≥ 3, or similar results among other tests for the assessment of cognitive impairment).
Literature reviews, meta-analyses, case reports, editorials, book comments and chapters, articles not in English language were excluded from this review work.
For each study, the following information was tabulated: author and year of publication, sample size, demographic data (mean age, gender), dropout rate, type of study design (RCT: Randomized Controlled Trial; QED: Quasi Experimental Design; NE: Non-experimental), residential care type, measures for dementia, measures for depression, intervention performed (type of intervention, qualification of the intervention provider, setting,

NOS 6
Frontiers in Public Health | www.frontiersin.org monotherapy/augmentation), main findings. Finally, the quality of each RCT was scored using the Jadad modified scale, 1 QED and NE were scored on the basis of the Newcastle-Ottawa Scale (NOS) (see text footnote 1). The Jadad modified (maximum 8 points) is a scale which assess the quality of studies (randomization, double-blinding, withdrawals and dropouts, inclusion and exclusion criteria, statistical analyses used) (see text footnote 1). The NOS (maximum 9 points) is a scale used for assessing the quality of QES and NE evaluating selection and comparability of cohorts and assessment of outcome (24). In order to estimate the effect size of each single intervention, we calculated standardized mean differences (SMDs) with 95% CIs to assess depression severity across different depression scales. We used Hedges' adjusted g statistic to calculate SMDs (it is very similar to Cohen's d statistic, but includes an adjustment for small sample bias). Standard mean differences were calculated using RevMan 5.4.

Study Selection and Characteristics
According to the inclusion/exclusion criteria described above, the review eventually included 56 full text articles; according to the type of intervention, studies were grouped in the following areas: horticulture/gardening (n = 3), pet therapy (n = 4), physical exercise (n = 9), psychoeducation/rehabilitation (n = 15), psychotherapy (n = 3), reminiscence and story sharing (n = 14), miscellaneous (n = 8, for those studies which did not seem to match any of the previous categories) (See the Synthesized findings section below for more details).
Sample size varied from 5 to 403 participants (we decided not to include case studies with <5 participants) and eligibility criteria were highly heterogeneous as well. Regarding methodological approach, 93% (n = 52) of the studies examined by this review were either Randomized Controlled Trials (RCT, n = 30), or had a Quasi-Experimental Design (QED, n = 22); 4 studies were had a Non-Experimental design (NE).
Tables 1-7 summarize the main features of the studies included in the review process: sample (Number-intervention group IG control group CG, Age, Gender W, Drop Out rate), study design, residential type, measures for dementia, measures for depression, intervention (-type, -individual or group setting, -professionals involved, -monotherapy or augmentation), duration and timing of treatment, main findings and quality rate (Jadad and NOS). We grouped studies according to intervention type as follows: horticulture/gardening, pets, physical exercise, psychoeducative/rehabilitation interventions, psychotherapy, reminiscence and story sharing, miscellaneous).

Horticulture and Gardening
Three studies on horticulture/gardening were included in our systematic review ( Table 1). These focused both on physical wellbeing and on the effects of increased social interaction involved in the activity itself. One study (21) described a significant decrease (i.e., an improvement) in the mean scores for depression in the experimental group, and the opposite in the control group (SMD −15.21; CI 95% −16.98 to −13.44). The randomized controlled trial by Lai et al. (22) highlighted a significant effect of the intervention from a subjective perspective (increased subjective happiness), but not from an objective one (GDS score). Last, the study by Park et al. (23) failed to observe any change in depression scores in elderly women living in Korean nursing homes, and a worsening of depression scores (from normal scores to moderate depression) in the control group during the observation period. SMDs, CI 95% are reported in Supplementary Table 1 and Supplementary Figure 1.

Pet Therapy
Four studies on pet therapy were included in our systematic review ( Table 2); pets were dogs in 3 of them and cats in one (38). Three of these studies (38)(39)(40) showed an improvement of elderly patients' depressive symptoms, while no significant change in the GDS score was found in Phelps' sample (5 patients) (41). With more detail, Ambrosi et al. (40) highlighted that dog assisted therapy could reduce depressive symptoms in elderly people by facilitation of social interactions and consequent elicitation of a positive emotional response (SMD −1.39; CI 95% −2.22

Physical Exercise
Nine studies dealt with physical exercise, including different approaches and exercise programmes ( Table 3). As reported in Supplementary Table 3 (61), mobility, strengthening, closed-kinetic chain, walk, and relaxation (25).
On the other hand, the effectiveness of Wii Fit (44) or Qui gong exercise (45) remain uncertain on participants' level of depression.

Psychoeducational and Rehabilitation Interventions
Fifteen studies on psychoeducational and rehabilitation interventions were included in our systematic review ( Table 4). The interventions in this group were performed in individual or group setting, either self-managed after training or managed by experienced staff. Among the heterogeneous interventions described, several authors reported a positive effect on depressive symptoms (26-33, 46, 62, 66, 71, 72). The types of psychoeducational/rehabilitation interventions assessed included: 8 week sessions of writing group (45); training, education and activity programmes for residents as well as for physicians and carers (26); a mixed intervention including different activities with a recreation therapist (27); an individualized psychosocial intervention carried out by a volunteer and a nurse (28); mentoring English conversational activity (29); self-worth therapy managed by a trained mental health nurse (30); a program of pleasant event activities (31); videoconference interaction with relatives (32); "friendly visits" problem-solving treatment (72); a specific multidisciplinary care program (66); a Socially Supportive Activities Program intervention (62); facility adaptation promotion program (46); nurse-led sleep program (33 Figure 4. Two out of these 15 studies failed to find significant changes in depressive symptoms: the study by Dozeman et al. (67), about a guided self-help course, and that by Cesetti et al. (47), dealing with a narrative group intervention.

Psychotherapy
Three studies on cognitive-behavior psychotherapy were included in the review ( Table 5). Two of them supported its effectiveness on depression (34,48), while one did not (77). SMDs for each study arm are reported in Supplementary Table 5 and Supplementary Figure 5.

Reminiscence and Story Sharing
Our review included 14 studies on reminiscence and story sharing ( Table 6). Reminiscence, life review and story sharing yielded significant results on depression in 12 studies (35-37, 49-54, 63, 73, 74) while two recent RCTs (68,75) failed to find a significant effect of the reminiscence therapy intervention on depressive symptoms.
SMDs are reported in Supplementary Table 6 and Supplementary Figure 6.

Miscellaneous
In the miscellaneous group we included 8 studies about different and heterogeneous approaches which did not seem to fit any of the above-described categories ( Table 7).
These miscellaneous interventions were: a pilot study regarding wheel prayer group activity (76); gero-transcendence, a particular form of transcendence intervention (55); Han et al.'s laugh protocol (56), a 1 month intervention program of transcutaneous electrical acupoint stimulation (TEAS) (57); a group activity based on ICT (using app for instant communications on electronic devices) Chiu and Wu (58); music therapy (59,64).
One study focused on the cost-effectiveness of a stepped-care program to prevent depression and anxiety in elderly homes (69).
All these miscellaneous interventions yielded some results, except for the one by Biasutti and Mangiacotti (59) and that by Bosmans et al. (69), which was neither cost-effective nor successful in preventing depression.
SMDs are reported in Supplementary Table 7 and Supplementary Figure 7.

Study Quality Assessment
The overall quality of RCTs was assessed with the Jadad; QED and NE studies were assessed using NOS and resulted quite poor (some studies involving small samples and/or with a poor methodological approach). The results regarding studies quality were consistent with those reported by the previous review by Simning and Simons (9) which highlighted an overall lack of quality of the studies in the field (using the Jadad scale).

Summary of Main Findings
The aim of this systematic review was to assess interventions focused on depression in elderly with no or mild cognitive impairment living in LTCFs. To our knowledge, three previous literature reviews (8,9,78) focused on depression among older people living in LCTFs and Nursing Homes and were performed with an approach which was similar to our own (at least as far as inclusion criteria and methods are concerned), albeit not totally overlapping. Simning and Simons (9) and Yoon et al. (8) included studies on both pharmacological and nonpharmacological interventions (psychotherapeutic, recreational, psychosocial and pharmacologic or other biologic interventions). Davison et al. (78) focused on psychological treatments in older adults living in LTCFs, but included those with a diagnosis of dementia or other comorbidities as well. Furthermore, they excluded rehabilitation and exercise-related activities and limited their review to reminiscence and life review interventions. Hence these review works eventually included different types of studies.
Studies included in the current review were highly heterogeneous as regards type, intervention methodology (duration, sessions number and frequency, setting -individual or group), number of participants, eligibility criteria, professional performing the intervention (members of the LTCF internal staff or external expert), type of residential facility.
Furthermore, as already reported by previous reviews (9,78), regrettably the studies available in the literature define depression criteria to a limited degree only. They usually lack a clinical diagnosis on behalf of an experienced psychiatrist; furthermore, most of the studies included in our review did not consider lifetime history of depression, comorbid medical illnesses or concomitant psychopharmacological treatments.
Most studies consider the decrease in depressive symptomatology level (investigated by validated psychometric questionnaires) as study outcome. Nonetheless, the assessment of changes in a depression rating scale score could not provide enough information about real mental health and clinical depression before and after intervention. The assessment of an actual improvement or remission of depression would imply a psychiatric interview and clinical instruments for diagnosis according to DSM or ICD criteria (78); actually, a clinical psychiatric assessment was one of the activities proposed by the stepped care approach in Leontjevas et al. (66).
The overall duration (weeks) of the intervention delivered seemed to have a greater impact on treatment effectiveness than the duration of each single session.
Forty-one percent of the studies considered in our systematic review were performed in the Asian continent; we can hypothesize that the rate of older adults in a given country, its cultural context, the perception of and the approach to aging might play a role from this standpoint. All these issues may eventually lead to differences in the attention paid to different types of non-pharmacological treatment for depressed elderly. For example, horticultures interventions were conducted only in the Asian continent, while most (75%) of the articles dealing with pet therapy come from Italy (38)(39)(40). This result can be related to the more holistic and spiritual vision of the Asian culture, characterized by a thorough approach including body, mind and spirit (79) and spirituality (80), which can increase Asian researchers' interest in non-pharmacological therapeutic approaches for elderly's depression.
Summarizing the main findings from the current review, most of the included studies seemed to suggest some kind of effectiveness of the proposed intervention on depression. Through our systematic literature review we were able to identify seven types of intervention, and in each one there are effective interventions with different effect sizes. Nonetheless, variability in sample size and eligibility criteria makes it difficult to compare results from different studies and to draw clear conclusions. Differences in participants' mean age, and even more, the choice to include/exclude soft or mild cognitive impairment, certainly affected the research findings. Due to this kind of heterogeneity a meta-analysis could not be performed; nonetheless, physical exercise, psychotherapy, and reminiscence interventions (see Supplementary Tables 3, 5, 6 and Supplementary Figures 3, 5, 6, respectively), seem to be the most effective interventions.
Furthermore, as described also by Simning and Simons (9), LTCFs constitute a widely heterogeneous context, spanning from low-level dependency to high-level dependency facilities, and hence offering different types of assistance. Regarding the populations involved in the studies we assessed, one more methodological limitation could be participants' selection: in many studies the sample was randomized, while in other studies sampling was largely dependent on type of LTCF, setting, population and residents' willingness to participate to the proposed intervention.
While the Yoon and co-workers' (8) review emphasized the importance of providing individualized treatments, no specific focus on individualization emerged from the articles included in the current review. Another limitation, which is inherent the specific population and setting we investigated, was about the frequency of drop out and the possibility for participants to complete the follow-up. Thus, the results should be considered with caution, in the light of the several limitations which are further detailed below.

Limitations
Some limitations of the current review should be underscored. We selected articles in English language only, which can potentially exclude relevant results published in other languages. We decided to focus on depressed elderly and excluded studies involving those with a severe cognitive impairment (MMSE score < 20), which represent only a portion of the global elderly population. Despite mixed or negative findings in some cases, most studies included in this systematic review reported that the non-pharmacological interventions they assessed seemed effective in the management of depressed elderly in the LTCFs context. Nonetheless, making a clear decision about their actual effectiveness is difficult as data reported are very varied, and the populations involved in the studies are varied as well, and sometimes poorly defined. Regrettably, this fact together with the limitations and heterogeneity of the studies described above, hinder the possibility to draw clear conclusions about the actual effectiveness of the interventions analyzed in the papers (78). Despite the limitations described above, we believe that the current review adds to the existing literature as it tried to focus as much as possible on not severely cognitively impaired depressed elderly, selecting studies involving participants with a MMSE score > 20; furthermore, it proposes a categorization of the types of non-pharmacological treatment offered in the context of LTCFs, and includes an accurate analysis of studies quality and effect size of the results obtained.

CONCLUSIONS
Non-pharmacological interventions for the management and treatment of depression in the elderly population could be an important strategy, as this population could be particularly frail and sensitive with regard to pharmacological interventions as a consequence of comorbid medical problems, concomitant drug therapies, different drug metabolism, higher probability of side effects.
Nonetheless, as described in the limitations section, the possibility to generalize results or to draw clear conclusions about the effectiveness of these interventions is limited. Further methodologically sound studies paying more attention on eligibility criteria, diagnosis, history comorbidity and medications are required to better understand whether a specific type of non-pharmacological intervention should be recommended for depressed elderly in LTCFs.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.