http://informahealthcare.com/dre ISSN 0963-8288 print/ISSN 1464-5165 online Disabil Rehabil, Early Online: 1–10 ! 2014 Informa UK Ltd. DOI: 10.3109/09638288.2014.932441

REVIEW PAPER

Farm-based interventions for people with mental disorders: a systematic review of literature Sorana C. Iancu1, Adriaan W. Hoogendoorn2, Marjolein B. M. Zweekhorst1, Dick J. Veltman2, Joske F. G. Bunders1, and Anton J. L. M. van Balkom2 Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

1

ATHENA Institute, Department of Innovation in the Health and Life Sciences, VU University Amsterdam, the Netherlands and 2Department of Psychiatry and EMGO+ Institute, VU University Medical Centre and GGZinGeest, Amsterdam, the Netherlands Abstract

Keywords

Purpose: Farms are increasingly used in mental healthcare. This study aimed to systematically review the evidence on the effectiveness of farm-based interventions for patients with mental disorders. Methods: Controlled and uncontrolled studies of farm-based interventions were included. Within- and between group effect sizes were calculated. Qualitative data were summarized using thematic synthesis. The review followed the PRISMA, Cochrane and COREQ standards. Results: The eleven articles included reported results of five studies, three of which were randomized control trials (RCTs). Overall, 223 patients with depressive disorders, schizophrenia or heterogeneous mental disorders attended three types of farms-based interventions. Favourable effects on clinical status variables were found in one study in patients with depressive disorders that did not respond to medication and/or psychotherapy, and in one RCT in patients with schizophrenia. Assessment of rehabilitative effects (functioning and quality of life) was limited and yielded conflicting results. Patients’ experiences revealed that social and occupational components of interventions were perceived as beneficial, and provided insights into how farm-based interventions may facilitate recovery. Conclusions: Our results suggest that the farm environment should be considered, especially for patients with mental disorders who do not achieve an adequate response with other treatment options. Further research is needed to clarify potential social and occupational benefits.

Depression, farm-based interventions mental disorders, schizophrenia History Received 7 January 2013 Revised 12 May 2014 Accepted 4 June 2014 Published online 25 June 2014

ä Implications for Rehabilitation   

Despite the developments in mental healthcare, in many countries farms still play a role in the provision of psychiatric rehabilitation services. Farm-based interventions can alleviate psychiatric symptoms in patients with persistent mental disorders and can facilitate mental health recovery. The social and occupational aspects of the farm-based interventions are central to the experiences of mental health recovery.

Introduction Rural spaces have long been considered to be places for recovery and rehabilitation but their use in mental healthcare has fallen in and out of favour during the past century. Historically, care of the mentally ill was provided in residential institutions located in remote areas, surrounded by farms and gardens. In the absence of other therapeutic options, farms and gardens were used for their restorative effects and as means of occupying the time of patients [1,2]. From the 1950s onwards, when psychoactive drugs became available, the model of the mental asylum was largely abandoned.

Address for correspondence: Sorana Iancu, ATHENA Institute, Faculty of Earth and Life Sciences (FALW), VU University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, the Netherlands. Tel: +31205986994. Fax: +31205987027. E-mail: [email protected]

Buildings were either transformed into mental health hospitals and outpatient clinics, or were dismantled, and the surrounding land was sold [3]. Mental healthcare services were being modernized, and farms no longer seemed relevant to the new developments in healthcare. By the 1980s, it became apparent that psychiatric treatments could not, by themselves, provide full mental health recovery. Increasing acknowledgement that mental disorders affected a range of life domains led to the development of the field of psychiatric rehabilitation. Using a series of psycho-social interventions, psychiatric rehabilitation aimed to assist people with mental disorders in developing the skills needed for living independent lives in the community [4,5]. Since an important goal of rehabilitation is represented by employment, a range of vocational services was developed. Initially, farm-based services were considered an interesting, though exceptional location for rehabilitation services [6]. Since the beginning of the century,

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

2

S. C. Iancu et al.

however, private farms in Western Europe and in the United States became involved in providing services for people with (mental) disabilities [7–10]. This development occurred as part of a broader phenomenon known in the literature as ‘‘green care’’, which also includes gardening, landscape or nature conservation, animal husbandry and exercise in natural environments [9]. Initial research on the potential benefits of farm-based services for people with mental disorders reported positive findings, such as significant improvements in mood and self-esteem [11], and self-reported feelings of social inclusion and empowerment [12,13]. The farm environment was also evaluated positively by mental healthcare professionals, who viewed the real-life setting of farms and the focus on users’ abilities as additional benefits to regular services [13] and expected that work with farm animals would improve participants’ social skills [14]. Despite these findings, to date no study reviewed systematically the effectiveness of farm-based interventions for people with mental disorders. This leaves unanswered questions about whether or not these interventions bring any added value to the current models of mental care and, if so, for which groups and under which circumstances. Answers to these questions could help mental healthcare service users and professionals make better choices among available services, and could guide public health and health policy makers in reconciling the increasing costs of mental healthcare with the limited resources available. In this context, the current study aimed to systematically review the evidence on the effectiveness of farm-based interventions for people with mental disorders.

Methods This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses standard (PRISMA) [15]. Identification of studies Relevant studies were identified through a search of the English language literature published before May 2014 in the electronic databases PsycINFO, MEDLINE and ISI Web of Knowledge. We posed no restrictions on the literature search regarding recency. Keywords were selected from the green care terminology [9]: care farm, animal-assisted therapy, green exercise, horticultural therapy, therapeutical horticulture, gardening, green care, in combination with farm. The reference lists of relevant articles were searched manually (forward citation tracking). In order to identify potentially relevant reviews on green care, the search was extended to the Cochrane Library. Inclusion and exclusion criteria Studies were included in the review if they met the following criteria: (i) evaluated farm-based interventions, defined as any type of green care activity (i.e. agriculture, gardening, animal keeping and husbandry) taking place on a farm; (ii) were designed as controlled or uncontrolled treatment outcome studies; (iii) included people with mental disorders, 18 years of age or older; (iv) used validated measurement scales (self-ratings or assessor ratings) to report outcomes related to mental health outcomes, defined as clinical status variables, functioning or quality of life. Interventions for people with mental retardation were excluded.

Disabil Rehabil, Early Online: 1–10

Data extraction Relevant data were identified using a coding scheme with codes for: article characteristics (year of publication and journal), participant recruitment (sampling method, inclusion/exclusion criteria, number of participants, number of completers and dropouts, reason for non-participation), intervention characteristics (type, duration and schedule, providers, other treatments received), participant characteristics (diagnosis, diagnostic criteria, age, sex), outcome measurements (timing of measurements, outcome measures, results on outcome measures of interest), and results (of quantitative and qualitative analysis). Data was extracted by the first author and checked by the third author. Statistical analysis Inclusion of both controlled and uncontrolled studies required the calculation of two different types of effect sizes [16]. Changes in mental health outcomes from the start to the end of the intervention (within-group effect size), were assessed using the standardized mean difference between pre- and post-test scores, corrected for small samples (Hedges’ g for repeated measures, gRM) [17]. In the controlled studies, participants in the experimental groups received the farm-based interventions and treatment as usual (TAU) or vocational rehabilitation (VR), whereas participants in the control groups received TAU or VR. In order to assess whether the addition of the farm-based intervention, the TAU or VR had an added value for the participants in the experimental group, we calculated between-group effect sizes which compare pre-post test changes in mental health outcomes in experimental groups to those in controls. Between-group effect sizes were calculated using the standardized mean difference between the pre-post test changes reported in the two groups, corrected for small samples (Hedges’ g for independent groups, gIG). The calculation of gRM and gIG was based on the standard deviation presented by authors in the original papers. Analysis was performed using Comprehensive Meta-Analysis, version 2.2.64 (Englewood, NJ). We performed both completerand intention-to-treat (ITT) analyses, assuming that scores would not change in participants who dropped out. Statistical significance was set at p  0.5. Magnitude of effect size estimates was interpreted as small (g ¼ 0.2), medium (g ¼ 0.5), or large (g ¼ 0.8) [18]. Given the clinical heterogeneity of the studies included in the review, no summary indicators were calculated. Thematic synthesis of participants’ experiences with farm-based interventions Data on participants’ experiences with the farm-based interventions, collected in three studies [19–22], were summarized through a three-stage thematic synthesis [23]. First, the main concepts identified by the authors of the original articles were extracted and summarized across studies. Thereafter, concepts were grouped into descriptive themes which were checked against the results of original articles in order to ensure that they remained true to the original data. The descriptive themes were then organized within analytical categories. The first stage was conducted independently by the first and third authors, who discussed possible interpretations of data. The second and third stages were completed by the first author, and checked by the third author. Divergences were solved through consensus, by referring to the original articles. Quality assessment The internal validity of the three RCT studies included in the review [21,22,24–28] was assessed across eight domains

Farm-based interventions for mental disorders

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

DOI: 10.3109/09638288.2014.932441

(sequence generation; allocation concealment; blinding of participants, investigators and outcome assessors; incomplete data reporting; selective outcome reporting; and other biases), in accordance to the Cochrane Handbook for Systematic Reviews of Interventions [29]. All RCTs included small samples; only one reported data on compliance [27]. Overall, we found a high-risk for performance bias (impossibility to blind participants and study personnel to experimental condition), and low-risk for selection bias (adequate allocation sequence, adequate concealment), attrition bias (overall drop-out rate of 17.1%) and reporting bias (no selective outcome reporting). Risk of detection bias was unclear, as outcome assessors were blinded only in one study [21]. We found no evidence for publication bias (symmetrical funnel plots). Data on participants’ experiences with farm-based interventions were collected in three studies [19–22]. In-depth interviews were conducted in two studies [21,22], but only one [22] included participant quotes and distinguished between major and minor themes, as recommended by the Consolidated Criteria for Reporting Qualitative Research (COREQ) [30]. The third study used questionnaires with closed- and open-ended items to collect data on participants’ experiences, and summarized results as Figure 1. Flow chart of systematic review process.

3

percentages [19] and as recurrent themes and statements [20] respectively.

Results Study selection The study selection process is summarized in Figure 1. The search of electronic databases generated 738 articles. These were screened on the basis of title and abstract; potentially relevant articles were reviewed in full text and checked against the inclusion criteria. Of the 738 citations, 72 were duplicates, 625 did not meet the eligibility criteria when screened on the basis of title and abstract, and 32 did not meet the eligibility criteria when examined in full text. As a consequence, 9 articles were included in the review. Forward citation tracking generated 20 additional articles, of which two met the eligibility criteria. No other relevant articles were identified using the Cochrane Library. The final sample included in the review consisted therefore of 11 articles reporting the results of 5 studies that were published between 2007 and 2014.

4

S. C. Iancu et al.

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

Description of studies Among the 5 studies, three were RCTs [21,22,24–28] and two did not use a control group [19,20,31–33]. In total, 223 participants were recruited (167 in experimental groups and 58 in control groups). The samples consisted of participants with depressive disorder [19,20,22,27,28,32,33], schizophrenia [21,31], or a variety of common and severe mental disorders [24–26]. Diagnosis was established by means of a structured clinical interview (namely, the MINI International Neuropsychiatric Interview, [34]) in two of the five studies [19,20, 22,27,28,32,33]. Farm-based interventions consisted of farm animal-assisted activities [22,24–28], horticulture [19–21,32,33] or therapeutic horse riding [31]. Interventions were evaluated in relation to clinical status [21,25,27,31–33], cognitive functioning [32,33], psychosocial functioning [19–21,24,27], occupational functioning [21,26] and quality of life [21,26]. Participants also provided data on their experiences with the farm-based interventions [19–22]. The aims, methods and results of the five studies are reviewed below. Farm-based interventions for participants with depressive disorders Farm-based interventions were evaluated as a form of complementary treatment for participants with depressive disorders in two studies from Norway [19,20,22,27,28,32,33]. Building on theories of social support and self-esteem, Pedersen et al [22,27,28] assessed the effects of a farm-animal assisted intervention consisting of work and contact with farm animals. Thirtyfour participants with major depressive disorder (79.3% women; mean age 37.8 years) receiving TAU (treatment as usual: medication and/or psychotherapy) were randomly assigned by a computer program to either the experimental or waiting list control group. The intervention was organized twice a week, for 12 weeks, on 11 farms. Sixteen participants in the experimental group and 13 in the control group started the intervention, but only 12 completed it in each group. Assessments were performed twice at baseline, at 4 and 8 weeks, at the end of intervention, and at 3 month follow-up. Assessors were not blinded. Outcomes included clinical status (severity of depressive and anxiety symptoms, measured with the Beck Depression Inventory, BDI [35], and the state version of the State-Trait Anxiety Inventory, STAI-SS [36] respectively), and psychosocial functioning (general self-efficacy, measured with the General Self-Efficacy Scale, GSE [37]). In 14 participants present in the experimental group, video recordings were used to estimate the time spent in various tasks early and late in the intervention. Experiences with the intervention were explored through in-depth interviews conducted with eight completers. The intention-to-treat analysis revealed significant improvements in the severity of depressive symptoms and general selfefficacy in the experimental group, but no differences were found when compared to controls [27]. Normal levels were achieved by 6 completers in the experimental group and by one in the control group. Severity of anxiety symptoms was not affected significantly [27]. The time spent in work activities increased during the intervention, and progressive involvement in complex tasks correlated significantly with changes in mental health outcomes [28]. Participants perceived the intervention as positive, due to participation in ordinary (work) life, distraction from disorders, flexibility to accommodate different levels of functioning, and experiences of coping [22]. Gonzalez et al. [19,20,32,33] evaluated the effects of a horticultural intervention in participants with depressive disorder. Forty-six participants (78.3% women; mean age 46.3 years) receiving TAU (medication and/or psychotherapy) attended twice

Disabil Rehabil, Early Online: 1–10

a week during three hours, for 12 weeks, gardening activities on four urban farms. No control group was used. Five participants dropped out early in the intervention. Assessments were performed twice at baseline, at 4, 8 and 12 weeks and at 3 month follow-up. Outcomes included clinical status (severity of anxiety and depressive symptoms assessed with BDI and STAI-SS respectively), cognitive functioning (perceived attentional capacity and rumination, assessed with Attentional Functional Index [38] and Ruminative Response Scale – Brooding Subscale [39] respectively), and psychosocial functioning (perceived Stress Scale [40], Positive and Negative Affect Scale (PANAS-PA) [41], Life Regard Index-Revised [42] and Sense of Coherence Scale (SOC) [43]). Existential issues were explored using open and close-ended questions. Analysis, based on completer data, showed that depression scores were stable between the two baseline measurements (when participants received only TAU), and declined in the first month of the intervention, suggesting that improvements were related to the intervention, rather than the associated use of TAU [32,33]. Significant improvements were also found in severity of anxiety symptoms, positive affect, perceived stress, rumination, attentional capacity and social activity [19,32], but not in existential issues [20]. Improvements in the severity of depressive symptoms were maintained at 3 month follow-up [32,33]. Farm-based interventions for participants with schizophrenia Two studies from China and Italy evaluated farm-based interventions for their potential for psychiatric rehabilitation in participants with schizophrenia [21,31]. In China, Kam and Siu [21] assessed the effects of a standardized horticultural intervention, using a vulnerabilitystress-coping perspective on mental disorders. Twenty-four participants (70.8% men, mean age: 44.3 years, 91.7% diagnosed with schizophrenia) were randomly assigned to the experimental group receiving the intervention and vocational rehabilitation (VR), or to the control group receiving VR only. The intervention consisted of one-hour structured horticultural sessions over 10 days, provided by an occupational therapist. VR consisted of unstructured horticulture activities and indoor industrial activities. Use of psychiatric treatments was not reported. Ten participants started the intervention in the experimental group and 12 in the control group. No further drop-outs occurred. Participants were assessed blindly before and after the intervention. Outcomes included clinical (severity of anxiety and depressive symptoms) and psychosocial functioning (levels of perceived stress), assessed with the Depression Anxiety Stress Scale [44]; occupational functioning (work behaviour, assessed with the Work Behaviour Assessment Scale [45]); and quality of life, assessed with the Personal Well-being Index [46]. No schizophrenia assessments were included. The completer analysis revealed that depression and anxiety scores improved significantly in the experimental group, compared to controls, but occupational functioning and quality of life were not affected. Interviews with participants in the experimental group suggested that the intervention was experienced as beneficial, helped enhance occupational performance and promote emotional health and social performance, and gave a feeling of connecting with nature [21]. In Italy, Cerino et al. [31] assessed the effects of therapeutic horse riding as a rehabilitation programme for patients with schizophrenia. Twenty-four male and female participants, aged 18 to 40 years, were included in an experimental group. No drop-out rates were reported. The intervention, held on 6 farm-based riding centres, consisted of 40 weekly, standardized sessions aiming to improve specific riding abilities, self-esteem and social functioning, and to reduce stigma. Use of psychiatric treatments was not

Farm-based interventions for mental disorders

DOI: 10.3109/09638288.2014.932441

reported. Assessments were performed before and after the intervention, and investigated changes in clinical status variables (negative and positive symptoms), using the Brief Psychiatric Rating Scale [47] and the Positive and Negative Syndrome Scale [48]. Results indicated significant improvements in symptoms, with most apparent changes across negative symptoms [31].

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

Farm-based interventions for participants with common and severe mental disorders Berget et al. evaluated a non-standardized, farm-animal assisted intervention both as a form of treatment, additional to psychiatric treatment [24,25], and as a form of occupational therapy [26]. The intervention, provided on 15 farms in Norway, was organized twice a week for 12 weeks. Ninety participants with ICD10 diagnoses of schizophrenia (35.3%), personality disorders (27.9%), depressive disorders (26.5%) and anxiety disorders (10.3%) were randomized using a computer program to either experimental group receiving the intervention and TAU (psychiatric medication), or to TAU control. Sixty-nine participants (75.6% women, average age 34.7 years) completed the intervention: 41 in the experimental group and 28 in the control group. Drop-out was related to disorder severity. Measurements, performed before and after the intervention, and at 6 month follow-up, assessed clinical status (severity of depressive and anxiety symptoms, measured with BDI and STAI-SS); psychosocial functioning (general self-efficacy, using GSE, and coping skills, using the Pressure Management Indicator-Coping Strategies Scale [49]); and quality of life (using the Quality of Life Scale [50]). In a subsample of 35 participants (16 completers), working abilities (intensity and exactness) were assessed early and late in the intervention, using video-recordings. Analysis, performed on completer data, revealed no significant changes in clinical status variables, psychosocial functioning or quality of life during the intervention [24,25], but showed significant improvements in work abilities in participants in the experimental group [26]. However, changes in work abilities did not correlate with changes in clinical status variables, functioning or quality of life [26]. At 6 month follow-up, significant improvements were found in severity of anxiety symptoms [25] and general self-efficacy [24] in the experimental group, compared to controls. Coping skills also improved in the experimental group but differences relative to controls were not reported [24]. Summarizing the effects of farm-based interventions The effects of farm-based interventions were assessed based on two types of effects sizes: within-group effects (i.e. the changes in mental health outcomes from the start to the end of the intervention), calculated for all the studies included in the systematic review, and between-group effects (i.e. difference in pre-post test changes in mental health outcomes between experimental and control groups), calculated for controlled studies. In order to allow comparison of effects within each of the three different study populations included in the original articles, results are presented separately for participants with depressive disorders, for participants with schizophrenia and for participants with common and severe mental disorders. Within-group effects The within-group effect sizes calculated across studies are presented in Table 1. For participants with depressive disorders [19,27], we found moderate to large improvements in severity of depressive and anxiety symptoms in both experimental and control groups. This suggests that, during the intervention,

5

participants with depressive disorders experienced clinically relevant improvements in depressive and anxiety symptoms, regardless of whether they attended farm-based interventions and TAU, or TAU only. In the experimental groups, some aspects of cognitive and psychosocial functioning also improved moderately. For participants with schizophrenia, the addition of horticulture to vocational rehabilitation [21] was associated with large improvements in depressive and anxiety symptoms and in perceived stress in the exposed, but not in the control group, suggesting an increased efficiency of the combined approach. Work behaviour and quality of life were not affected, possibly due to the relatively short duration of the study (10 days). The oneyear therapeutic horse riding intervention [31] was associated with moderate to large improvements in positive and negative psychotic symptoms. Participants with common and severe mental disorders [24,25] experienced small and non-significant improvements in severity of depressive and anxiety symptoms, self-efficacy, coping and quality of life. This indicates that in heterogeneous samples, neither the addition of farm-based intervention to TAU, nor TAU alone induced improvements in clinical status, psychosocial functioning or quality of life. Between-group effects The between-group effects corroborated the results presented above. In both completer and intention-to-treat analyses (Tables 2 and 3, respectively), small and non-significant effects were found for depressive disorder [27], confirming that improvements in participants who received both the intervention and TAU were similar to those who received TAU only. In contrast, in participants with schizophrenia [21], effects on severity of anxiety symptoms, and on perceived stress, were significantly larger than for controls. Finally, in participants with common and severe mental disorders [24,25], the between-group analysis yielded small and non-significant effect sizes, confirming the similar trajectories of the experimental and control groups. Participants’ experiences with farm-based interventions The thematic synthesis of qualitative data identified three analytic categories of themes related to disability, recovery and specific farm experiences (Table 4). Referring to disability, participants described how the farm environment provided distraction, stress release and warm human and/or animal contact, thus alleviating suffering; how their limitations were accepted and taken into account on farms; and how flexible tasks were created to allow them to participate in work, despite their limitations. Recovery experiences were presented by participants who referred to a change in the way they viewed themselves because of others’ respect and appreciation, and because they realized that they could manage and accomplish tasks. Additionally, participants described that, on farms, they felt like ordinary workers (having an ordinary work life, improving or learning skills), they built up self-determination (improving motivation and feeling strengthened to continue during difficult periods), and felt socially included (increased social contacts, perceived improvements in social skills). Specific aspects of farm-based interventions referred to working with animals (farm-animal assisted interventions), enjoying the leisurelike activities, becoming absorbed in working with plants and being connected with nature (horticulture).

Discussion The primary aim of this study was to systematically review the evidence on the effectiveness of farm-based interventions for

6

S. C. Iancu et al.

Disabil Rehabil, Early Online: 1–10

Table 1. Within-group effect sizes Hedge’s g in mental health outcomes across the studies included in the review (completer analysis).

Depressive disorders

Study population

Pedersen1 [27,28]

First author References

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

Intervention groups Condition n per group Clinical status variables Severity of depressive symptoms Severity of anxiety symptoms Severity of schizophrenia symptoms BPRS PANNS Cognitive functioning Attentional capacity Rumination Psychosocial functioning Self-efficacy Stress Coping Positive affect Existential issues Occupational functioning Work behaviour Work intensity 2 Work exactness 2 Quality of life

Common and severe mental disorders

Schizophrenia

Gonzalez [19,20,32,33]

Kam [21]

Cerino [31]

Berget [24–26]

Experimental Control Experimental Experimental Control Experimental Experimental Control TAU F-AAI + TAU TRI VR HI + VR HI + TAU TAU F-AAI + TAU (n ¼ 28) (n ¼ 41) (n ¼ 24) (n ¼ 12) (n ¼ 10) (n ¼ 46) (n ¼ 12) (n ¼ 12) gRM

p

0.88 0.83

*** ***

– –

– –

– –

– –

– –

0.66 – – – – – – – –

gRM

p

gRM

p

gRM

0.83 0.45

*** ***

0.92 1.07

– –

– –

– –

– –

– –

– –

– –

– –

0.52 0.49

*** **

– –

– –

** – – – –

0.41 – – – –

ns – – – –

– 0.47 – 0.32 0.16

– *** – ns ns

– 0.77 – – –

– – – –

– – – –

– – – –

– – – –

– – – –

0.29 – – 0.04

0.69 ** 0.91 ***

p

gRM

p

gRM

p

gRM

p

– –

– –

0.21 0.14

ns ns

– –

0.84 0.64

*** ***

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– ** – – –

– 0.05 – – –

– ns – – –

– – – – –

– – – – –

0.06 – 0.14 – –

ns – ns – –

0.04 – 0.09 – –

ns – ns – –

ns – – ns

0.08 – – 0.09

ns – – ns

– – – –

– – – –

– 1.11 1.10 0.00

– *** *** ns

– – – 0.08

– – – ns

*** 0.13 ns *** 0.08 ns

gRM

p

0.18 ns 0.08 ns

Bold values indicate within-group effect sizes which reached statistical significance. RM ¼ repeated measurements; gRM reflects the difference from pre-test to post-test in each group: positive values indicate that scores improved at posttest, compared to pre-test. F-AAI: farm animal-assisted intervention; TAU ¼ treatment as usual (medication or psychotherapy); HI: horticultural intervention; VR ¼ vocational rehabilitation; TRI ¼ therapeutic riding intervention; 1 ¼ original article reported ITT data; calculations for this table were based on the assumption that scores would not change in drop-outs; 2 ¼ based on a subsample of 19 completers from the experimental group; **p  0.01, ***p  0.001, ns – not significant. Table 2. Between-group effect sizes Hedge’s g in mental health outcomes in the RCTs included in the review (completer analysis). Study population First author References Experimental versus control condition n (completers) Clinical status variables Severity of depressive symptoms Severity of anxiety symptoms Psychosocial functioning Self-efficacy Stress Coping Occupational functioning Work behaviour Quality of life

Depressive disorders Pedersen [27,28] F-AAI + TAU versus TAU (n = 24)

Schizophrenia Kam [21] HI + VR versus VR (n = 22)

Common & severe mental disorders Berget [24,25] F-AAI + TAU versus TAU (n = 69)

gIG

p

gIG

p

gIG

p

0.25 0.04

ns ns

0.89 1.27

* **

0.04 0.22

ns ns

0.11 – –

ns – –

– 0.86 –

– * –

0.11 – 0.23

ns – ns

– –

– –

0.26 0.08

ns ns

– 0.08

– ns

Bold values indicate between-group effect sizes which reached statistical significance. F-AAI: farm animal-assisted intervention; TAU ¼ treatment as usual (medication or psychotherapy); HI: horticultural intervention; VR ¼ vocational rehabilitation. IG ¼ independent groups, completer analysis; gIG reflects the difference between experimental and control groups: positive values indicate that the intervention favours participants in experimental group; *p  0.05, **p  0.01, ns – not significant.

adult participants with mental disorders. A systematic review of the English language literature published before May 2014 identified 11 articles reporting the results of five studies (three randomized control trials and two uncontrolled studies).

Studies approached the farm-based interventions from different theoretical perspectives, illustrating the complexity of the farm environment which can act as a natural place with restorative qualities [32,33], provide opportunities for work [21,26,27], and

Farm-based interventions for mental disorders

DOI: 10.3109/09638288.2014.932441

7

Table 3. Between-group effect sizes Hedge’s g in mental health outcomes in the RCTs included in the review (intention-to treat analysis, ITT). Study population First author References Experimental versus control condition n (intention-to-treat)

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

Clinical status variables Severity of depressive symptoms Severity of anxiety symptoms Psychosocial functioning Self-efficacy Stress Coping Occupational functioning Work behaviour Quality of life

Depressive disorders Pedersen [27,28] F-AAI + TAU versus TAU (n = 29)

Common & severe mental disorders Berget [24,25] F-AAI + TAU versus TAU (n = 90)

Schizophrenia Kam [21] HI + VR versus VR (n = 24)

gIG-ITT

p

gIG-ITT

p

gIG-ITT

p

0.05 0.07

ns ns

0.69 1.03

# *

0.02 0.17

ns ns

0.00 – –

ns – –

– 0.82 –

– * –

0.01 – 0.19

ns – ns

– –

– –

0.35 0.07

ns ns

– 0.07

– ns

Bold values indicate between-group effect sizes which reached statistical significance. F-AAI: farm animal-assisted intervention; TAU ¼ treatment as usual (medication or psychotherapy); HI: horticultural intervention; VR ¼ vocational rehabilitation, IG-ITT ¼ independent groups, intention-to treat analysis; gIG-ITT reflects the difference between experimental and control groups: positive values indicate that the intervention favours participants in experimental group. #p  0.1, *p  0.05, ns – not significant. Table 4. Participants’ experiences with farm-based interventions, in relation to disability, recovery and specific experiences. Study population First author References (n) Themes related to disability Setting accepting disability Intervention accommodating disability Intervention alleviating disability Themes related to recovery Positive appraisals from others (verbal or non-verbal) Becoming aware of one’s own potential Change in view of life Developing a worker identity Building up self-determination Feeling socially included Themes of specific experiences Pleasurable, absorbing activities Working with animals Connecting with nature

facilitate social support among peers [19] and from farmers [28]. However, the quality of the studies included in the review was moderate in our opinion, as some had small sample sizes, one used a measurement instrument better suited for people with depression disorder, rather than schizophrenia, one reported qualitative data succinctly, and none controlled for TAU. This probably conveys the difficulties in implementing interventions in the complex environments of farms, detached from the clinical practice. Furthermore, the different studies used different ways of reporting the results. In order to allow comparison of the effects across studies, we transformed results based on intention to treat analysis into completer analysis (Tables 1 and 2), assuming that scores in drop-outs would be similar to those in participants who completed the interventions. However, it is likely that attrition rates were higher among participants who had more severe depressive symptoms at the beginning of the intervention, or who perceived the intervention as less useful. Indeed, one of the study included in the review explicitly reported such findings [24].

Depressive disorders Pedersen [22] (n ¼ 8)

Gonzalez [19,20] (n ¼ 46)

ˇ ˇ ˇ

ˇ

ˇ ˇ

ˇ

Schizophrenia Kam [21] (n ¼ 10)

ˇ ˇ ˇ

ˇ ˇ ˇ ˇ

ˇ ˇ

ˇ ˇ ˇ

ˇ

ˇ

ˇ

ˇ

ˇ

This implies that, if the drop-outs would have continued with the intervention, the final depression scores would have been lower than the ones found in completers, making the estimate calculated for this study (namely, [27]) an overestimation of the true effect. However, since the between-group analysis yielded no significant differences between experimental and control groups, this potential bias had little influence on the overall conclusion of the systematic review. All studies assessed farm-based interventions as forms of treatments, either alone or complementary to routine psychiatric treatment. The results of the studies included in this review suggest promising first results for schizophrenia (i.e. positive and negative symptoms of [31] and comorbid anxiety symptoms [21]). Interestingly, in both controlled [27] and uncontrolled [32,33] studies involving participants with depressive disorders, farmbased interventions were also associated with large improvements in the severity of depressive symptoms. In the study of Pedersen [27], no statistically significant difference was found between the

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

8

S. C. Iancu et al.

experimental group (i.e. participants receiving intervention and TAU) and the control group (i.e. participants receiving TAU only). However, the effect size in the control group was large (namely, 0.69), suggesting that participants in this study responded well to TAU. The study of Gonzalez [32,33] included no control group. To compensate for this shortcoming in the study design, two baseline measurements were conducted before the intervention (i.e. when participants received TAU only). Between these measurements, depression scores remained stable, suggesting that participants in this group were resistant to TAU. Therefore, we expect that a control group with similar characteristics would have showed a low effect size for TAU. Since the depression scores decreased during the first month of the intervention, we would further expect a large difference in effect size between the control and the experimental groups. However, since no control group was in fact available in the study of Gonzalez, this scenario remains an informed guess. Nevertheless, its exploration is worthwhile, given the difficulties in treating the individuals with depressive disorders that do not respond to medication and/or psychotherapy. The relevance of farm-based interventions for recovery and rehabilitation was explored only superficially by the studies included in the current review. Although multiple domains of functioning were assessed (i.e. cognitive, psychosocial and occupational), important knowledge gaps persist in understanding how farm-based interventions might contribute to rehabilitation of people with mental disorders. One domain that needs further investigation is occupational functioning which provided conflicting results: although observations and in-depth interviews suggested that work skills improved [22,26,28], use of validated questionnaires yielded no effects, probably due to the short duration of the intervention (10 days) [21]. A second domain that requires further attention is social functioning, an outcome of major importance in psychiatric rehabilitation [5]. Previous qualitative studies had indicated that farm-based services could improve social functioning [13,14] but none of the studies we reviewed assessed potential changes in this domain. Remarkably, quantitative and qualitative data both seemed to suggest that the farm-based environments facilitated supportive social interactions: participants reported high levels of group cohesiveness [19], and described communication with farmers and peers as beneficial [21,22]. However, it remained unclear whether the perceived quality of the social component of farmbased interactions was related to the possibility to interact with others or to specific characteristics of the social environment on farms. The results of the thematic synthesis of participants’ experiences on farms brought forward a different perspective on how farm-based interventions might facilitate recovery. Reference to improvements in clinical status and functioning were largely absent from participants’ accounts, despite the fact that the studies in which they participated focused on such outcomes. The only notable exception was stress which improved both objectively [19,21] and subjectively [21]. In their accounts, participants referred to experiences of balancing disability and recovery, finding new meaning and redefining a positive sense of identity, and described learning new ways to deal with disability and discovering new ways to view oneself and life [21,22], much in agreement with the literature on mental health recovery [51–58]. A recent literature review described recovery as a multi-staged transition from being overwhelmed with disability, to ignition of hope and awareness of a better life, and learning to live with and beyond disability [59]. The qualitative results of our review, although based on limited data, seem to suggest that such processes also occur on farms. However, none of the studies explored how participants

Disabil Rehabil, Early Online: 1–10

experienced the farm-based interventions, compared to other interventions, despite the fact that participants in two studies also attended other forms of rehabilitation [21,22]. Insights into similarities, as well as differences, between processes of recovery on farms and other rehabilitation services could result in a better understanding of the added value of farms, and facilitate a better match between services and needs of people with mental disorders. To the best of our knowledge, the current research represents the first attempt to review systematically the literature on farm-based interventions for people with mental disorders. Our literature search yielded a wealth of qualitative and observational studies on various green care interventions, and for different participant groups, suggesting that the topic of green care attracts a great amount of interest from the academic community. However, the number of high-quality studies seems to be limited, despite the calls for better evidence for green care interventions, as expressed by practitioners, researchers and healthcare professionals [60]. The articles included in this review originated from three different countries, and were published within the last five years. This suggests that researchers are becoming aware of the importance of building stronger evidence for green care. We recommend that future research on farm-based interventions addresses the methodological issues we identified in the literature. In our view, subsequent studies should strive to use high-quality designs and to incorporate validated measures of functioning. For the latter, questionnaires that cover multiple domains – such as the World Health Organization Disability Assessment Schedule [61], WHO DAS II – could help increase the comparability of results across studies and populations. A last recommendation refers to the exploration of processes of recovery on farms, as compared to other rehabilitation services.

Declaration of interest This study was funded by TransForum Netherlands, a temporary collaborative programme between research institutes, Dutch government, non-governmental organisations and private sector, which conducted research on agriculture in a metropolitan context. TransForum Netherlands had no involvement in the design, data collection, analysis and interpretation, in the writing of the report or in the decision to submit the paper for publication. The authors report no conflicts of interest.

References 1. Fause A. Care for the mentally ill in the Norwegian counties Troms and Finnmark 1891–1940. Tidsskr Nor Laegeforen 2008;18: 2860–63. 2. Sullivan ME. Horticultural therapy – the role gardening plays in healing. J Am Health Care Assoc 1979;5:5–6. 3. Ravelli DP. Deinstitutionalisation of mental health care in The Netherlands: towards an integrative approach. Int J Integr Care 2006;6:e04–15. 4. Farkas M, Anthony WA. Psychiatric rehabilitation interventions: a review. Int Rev Psychiatry 2010;22:114–29. 5. Ro¨ssler W. Psychiatric rehabilitation today: an overview. World Psychiatry 2006;5:151–7. 6. Corbiere M, Lecomte T. Vocational services offered to people with severe mental illness. J Ment Health 2009;18:38–50. 7. Di Iacovo F, O’ Connor D. Supporting policies for social farming in Europe. Firenze: Arsia; 2009. 8. Hassink J, van Dijk M. Farming for health. Green-care farming across Europe and the United States of America. Wageningen, the Netherlands: Springer; 2006. 9. Haubenhofer DK, Elings M, Hassink J, Hine RE. The development of green care in Western European countries. Explore 2010;6: 106–11.

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

DOI: 10.3109/09638288.2014.932441

10. Sempik J, Hine R, Wilcox D. Green Care: A conceptual framework, a report of the working group on the health benefits of green care, COST Action 866, Green Care in Agriculture. Loughborough: Centre for Child and Family Research, Loughborough University; 2010. 11. Hine R, Peacock J, Pretty J. Care farming in the UK: Contexts, benefits and links with therapeutic communities. Int J Ther Communities 2008;29:245–60. 12. Elings M, Hassink J. Green care farms, a safe community between illness or addiction and the wider society. Int J Ther Communities 2008;29:310–21. 13. Hassink J, Elings M, Zweekhorst M, et al. Care farms in the Netherlands: attractive empowerment-oriented and strengths-based practices in the community. Health Place 2010;16:423–30. 14. Berget B, Ekeberg O, Braastad BO. Attitudes to animal-assisted therapy with farm animals among health staff and farmers. J Psychiatr Ment Health Nurs 2008;15:576–81. 15. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009;6:1–28. 16. Morris SB, Deshon RP. Combining effect size estimates in metaanalysis with repeated measures and independent-groups designs. Psychol Methods 2002;7:105–25. 17. Hedges L. Distribution theory for Glass’s estimator of effect size and related estimators. J Educ Behav Stat 1981;6:107–28. 18. Cohen J. Statistical power analysis for the behavioral sciences. New York: Erlbaum; 1998. 19. Gonzalez MT, Hartig T, Patil GG, et al. A prospective study of group cohesiveness in therapeutic horticulture for clinical depression. Int J Ment Health Nurs 2011;20:119–29. 20. Gonzalez MT, Hartig T, Patil GG, et al. A prospective study of existential issues in therapeutic horticulture for clinical depression. Issues Ment Health Nurs 2011;32:73–81. 21. Kam MCY, Siu AMH. Evaluation of a horticultural activity programme for persons with psychiatric illness. HKJOT 2010;20: 80–6. 22. Pedersen I, Ihlebaek C, Kirkevold M. Important elements in farm animal-assissted interventions for persons with clinical depression: a qualitative interview study. Disabil Rehabil 2012;34: 1526–34. 23. Thomas J, Harden A. Methods for the thematic synthesis of qualitative research in systematic reviews. BMC Med Res Methodol 2008;8:45. 24. Berget B, Ekeberg Ø, Braastad BO. Animal-assisted therapy with farm animals for persons with psychiatric disorders: effects on self-efficacy, coping ability and quality of life, a randomized controlled trial. Clin Pract Epidemiol Ment Health 2008;4:9–15. 25. Berget B, Ekeberg Ø, Pedersen I, Braastad BO. Animal-assisted therapy with farm animals for persons with psychiatric disorders: effects on anxiety and depression, a randomized controlled trial. Occup Ther Ment Health 2011;27:50–64. 26. Berget B, Skarsaune I, Ekeberg Ø, Braastad BO. Humans with mental disorders working with farm animals: a behavioral study. Occup Ther Ment Health 2007;23:101–17. 27. Pedersen I, Martinsen E, Berget B, Braastad BO. Farm animalassisted intervention for people with clinical depression: a randomized controlled trial. Anthrozoos 2012;25:149–60. 28. Pedersen I, Nordaunet T, Martinsen EW, et al. Farm animal-assisted intervention: relationship between work and contact with farm animals and change in depression, anxiety, and self-efficacy among persons with clinical depression. Issues Ment Health Nurs 2011;32: 493–500. 29. Higgins JPT, Altman DG. Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Grees S, eds. Cochrane handbook for systematic reviews of interventions. Version 5.0.1. Chichester, UK: John Willey & Sons, Ltd.; 2008:187–242. The Cochrane Collaboration. 30. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care 2007;19:349–57. 31. Cerino S, Cirulli F, Chiarotti F, Seripa S. Non-conventional psychiatric rehabilitation in schizophrenia using therapeutic riding: the FISE multicentre Pindar project. Ann Ist Super Sanita 2011;47: 409–14.

Farm-based interventions for mental disorders

9

32. Gonzalez MT, Harding T, Patil GG, et al. Therapeutic horticulture in clinical depression: a prospective study. Res Theory Nurs Pract 2009;23:312–28. 33. Gonzalez MT, Hartig T, Patil GG, et al. Therapeutic horticulture in clinical depression: a prospective study of active components. J Adv Nurs 2010;66:2002–13. 34. Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Norwegian version. J Clin Psychiatry 1998;59:22–33. 35. Beck AT, Steer RA. Manual for the beck depression inventory. San Antonio (TX): The Psychological Corporation; 1987. 36. Spielberger CD, Gorsuch RI, Lushene RE. Manual for the State-Trait Anxiety Inventory. Palo Alto (CA): Consulting Psychologists Press; 1983. 37. Schwarzer R, Jerusalem M. Generalized self-efficacy scale. In: Weinman J, Wright S, Johnston M, eds. Measures in health psychology: a user’s portfolio. Causal and control beliefs. Windsor, UK: NFER-NELSON; 1995:35–7. 38. Cimprich B. Development of an intervention to restore attention in cancer patients. Cancer Nurs 1993;16:83–92. 39. Treynor W, Gonzalez R, Nolen-Hoeksema S. Rumination reconsidered: a psychometric analysis. Cognitive Ther Res 2003;27: 247–59. 40. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983;24:385–96. 41. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affects – the PANAS scales. J Pers Soc Psychol 1988;54:1063–70. 42. Debats DL. Measurement of personal meaning: the psychometric properties of the life regard index. In: Wong PTP, Fry PS, eds. The human quest for meaning. A handbook of psychological research and clinical applications. Mahwah (NJ): Lawrence Erlbaum; 1998:237–60. 43. Antonovsky A. Unraveling the mystery of health: how people manage stress and stay well. San Francisco (CA): JosseyBass; 1987. 44. Henry JD, Crawford JR. The short-form version of the Depression Anxiety Stress Scales (DASS-21): construct validity and normative data in a large non-clinical sample. Br J Clin Psychol 2005;44: 227–39. 45. New Life Psychiatric Rehabilitation Association. Work behaviour assessment guidelines. Hong Kong; 2005. 46. Lau ALD, Cummins RA, McPherson W. An investigation into the cross-cultural equivalence of the Personal Wellbeing Index. Soc Indic Res 2005;72:403–30. 47. Overall J, Gorham D. The Brief psychiatric rating scale. Psychol Rep 1962;10:799–812. 48. Kay SR, Fiszbein A, Opler LA. The Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Schizophr Bull 1987; 13:261–7. 49. Cooper CL, Sloan SJ, Williams S. Occupational stress indicator. England: NFER-Nelson; 1988. 50. Wahl A, Burckhardt C, Wiklund I, Hanestad BR. The Norwegian version of the Quality of Life Scale (QOLS-N): a validation and reliability study in patients suffering from psoriasis. Scand J Caring Sci 1988;12:215–22. 51. Anthony WA. Recovery from mental illness: the guiding vision of the mental health service system in the 1990s. Psychiatr Rehabil J 1993;16:11–23. 52. Corrigan PW, Mueser TK, Bond GR, et al. Principles and practice of psychiatric rehabilitation. An empirical approach. New York: The Guildford Press; 2008. 53. Davidson L, O’Connell MJ, Tondora J, et al. Recovery in serious mental illness: a new wine or just a new bottle? Prof Psychol Res Pr 2005;36:480–7. 54. van Lith T, Fenner P, Schofield M. The lived experience of art making as a companion to the mental health recovery process. Disabil Rehabil 2011;33:652–60. 55. Onken SJ, Craig CM, Ridgway P, et al. An analysis of the definitions and elements of recovery: a review of the literature. Psychiatr Rehabil J 2007;31:9–22. 56. Repper J, Perkins R. Social inclusion and recovery: a model for mental health practice. London: Elsevier Science; 2003.

10

S. C. Iancu et al.

Disabil Rehabil Downloaded from informahealthcare.com by University of Laval on 07/02/14 For personal use only.

57. Roberts G, Wolfson P. The rediscovery of recovery: open to all. APT 2004;10:37–49. 58. Slade M, Amering M, Oades L. Recovery: an international perspective. Int J Epidemiol Psychiatric Sci 2008;17:128–37. 59. Leamy M, Bird V, Le Boutillier C, et al. Conceptual framework for personal recovery in mental health: systematic review and narrative synthesis. Br J Psychiatry 2011;199:445–52.

Disabil Rehabil, Early Online: 1–10

60. Sempik J, Aldridge J, Becker S. Social and therapeutic horticulture: evidence and messages from research. Reading: Thrive and Loughborough: CCFR; 2003. 61. Chwastiak LA, von Korff M. Disability in depression and back pain – evaluation of the World Health Organization Disability Assessment Schedule (WHO DAS II) in a primary care setting. J Clin Epidemiol 2003;56:507–14.