. 2022 Apr 14;19(8):4766. doi: 10.3390/ijerph19084766

Table A3.

Characteristics and findings of mixed methods studies included in review (n = 2).

Study (Author & Year)	Country/Setting	Place-Based Intervention (by Category)	Theory of Change/Likely Mechanisms	Sample Size and Characteristics (Total Size, % Female, Mean Age)	Means of Data Collection, Type of Data Collected	Analytic Approach	Key Findings (Effect Sizes, Key Themes, Efforts to Combine Findings from Quantitative and Qualitative Analysis)	Potential Harms Identified	Methodological Limitations
Active engagement in local green spaces
Gerber et al., 2017 [56]	United States; Bhutanese refuges recruited from local community garden and Bhutanese community network	Local community gardening at two local community plots. Authors clarified that these were in an urban area, and that some participants had to use the bus to access the gardens, but all were local to residents.	- Gardening can boost connectivity and community strengths particularly for groups highly valuing communal functioning and cohesion - Expected outcome: Bhutanese gardeners would self-repot significantly fewer symptoms of distress and perceive higher levels of social support	- 50 people (62% female (n = 31)) - mean age = 44.5, SD = 15.01 - 56% non-gardeners (n = 28), 44% gardeners (n = 22)	Quantitative data: Structured questionnaires to collect cross-sectional data on: - symptoms of PTSD, anxiety, depression assessed by Refugee Health Screener-15 (RHS-15) in Nepali - Perceived social support using Medical Outcomes Study Social Support Survey (MOS SSS) - Somatization using the Patient Health Questionnaire-15 (PHQ-15), a 15-item self-report scale of somatization derived from the Patient Health Questionnaire, used previously with refugee populations, and translated into Nepali - Functional adaptation using Adapted Client Assessment Tool (ACAT) to describe current mastery or independence in 13 domains of functioning, orally administered Qualitative data: Semi-structured interviews to explore social support issues (such as the nature of social interactions whilst gardening), local acculturation (such as degree of adjustment to life in the US), and perceived advantages and disadvantages of the community garden. Group meetings were held with participants to study findings and explore implications.	Descriptive analysis of quantitative data. Adapted form of thematic analysis of interview transcripts for 8 gardeners and 4 non-gardeners, using the approach of Consensual Qualitative Research (CQR), which involved both deductive and inductive methods to code and interpret data. Focus groups were then used to discuss findings of analyses of the data above to gain feedback on results and consider implications in a culturally salient way.	Quantitative: - RHS indicated that distress was not significantly different between gardeners and nongardeners - MOS SSS: gardeners reported greater social support (a moderate to large effect size (d = 0.70; 95% CI 0.12, 1.27), subscales revealed that gardeners reported significantly more tangible social support (large effect size, d = 0.88, 95% CI 0.28, 1.45), medium effect size was observed for emotional/informational social support (d = 0.53, 95% CI −0.04, 1.09), small to medium effect for affectionate support (d = 0.34, 95% CI −0.23, 0.89), and positive social interaction (d = 0.31, 95% CI −0.26, 0.86) - PHQ-15 scores found to be not statistically significant but found a small to medium effect size (d = 0.36, 95% CI −0.21, 0.91) with gardeners having more somatic complaints - Exploratory correlational analyses: Age was additionally significantly related to perceived emotional/informational social support (r = −0.38, p = 0.007; 95% CI −0.60, −0.11), with perceived support decreasing with age; ACAT was significantly related to time in US (r = 0.30, p = 0.033, 95% CI 0.02, 0.53) Qualitative: - Key themes: General findings (results presented in all, or all but one), typical findings (more than half of the cases), variant findings (2 to 4 cases)	None	- Self-selection to group membership leading to gardeners significantly more likely to live in a house and lower medical bills - Randomization of assignments may adversely disruptive to the ethos of the established community - Elderly with physical limitations may have difficulty obtaining certain benchmarks - Small to moderate effect sizes may be missed, although sample size was sufficient for hypothesis testing - Under-reporting of adjustment difficulties as participants feeling discomfort voicing their problems in interviews
Chiumento et al., 2018 [57]	United Kingdom; Children recruited from 3 schools in the North West of England (two primary schools and one secondary school)	Haven Green Space school garden project, involving monthly sessions over the course of 6 months in which schoolchildren were supported at school by two horticulturists and a Child and Adolescent Mental Health Service (CAMHS) psychotherapist to work together in designing a green space	Promote positive mental, emotional and physical wellbeing of the children with the “Five Ways to Wellbeing” framework (Connecting with others; Being active; Taking notice of the local environment and of their feelings; learning horticultural skills and how to manage successes and failure; Giving back to the wider community	36 children (14 females) - children with behavioural, emotional and social difficulties - 2 primary schools (year 5 and 6, aged 9–12) and 1 secondary school (year 7 to 9, aged 12–15)	Quantitative data: Collection of pre- and post- intervention scores on the following measures for children - Mental wellbeing using Wellbeing check cards (based on the 7-item version of the Warwick Edinburgh Mental Wellbeing Sale) Qualitative data: collected over the course of 2 h workshops (pre- and post- intervention) by using the Mental Wellbeing Impact Assessment (MWIA) to plot data in a participatory way with children under the following three domains: (1) enhancing control (2) increasing resilience and community assets (3) participation and social inclusion. This qualitative tool aims to assess the potential impact of a specific policy, service, project or program on the mental wellbeing of a population, and was originally developed for adults.	Quantitative analysis: Statistical comparison of scores on pre- and post- intervention measures. Qualitative analysis: thematic analysis of data, with the coding process deductively driven by the MWIA themes. Separate analyses of the quantitative and qualitative data then converged in the discussion to triangulate findings.	Quantitative: -wellbeing scores not found to be statistically significantly different from pre- to post- intervention (although no test statistics were presented to support this statement) Qualitative: Analysis of MWIA plots produced during workshops identified pre- and post-intervention tendencies towards pro-social behaviour (“feeling involved”, “having a valued role”, “sense of belonging” and “social networks and relationships”) and emotional symptoms. The thematic analysis also found that factors relating to mental health and wellbeing were positively impacted, including “emotional wellbeing” and “self-help”.	None	- Lack of a valid control group and small sample size - Lack of validation of Wellbeing check cards in children or of the adaptation of the MWIA for children (in view of its - use of adult terminology) - Wellbeing check cards: limited in their intended ability to capture change over time

Study
(Author & Year)

Country/Setting

Place-Based Intervention (by Category)

Theory of Change/Likely Mechanisms

Sample Size and Characteristics
(Total Size, % Female, Mean Age)

Means of Data Collection, Type of Data Collected

Analytic Approach

Key Findings (Effect Sizes, Key Themes, Efforts to Combine Findings from Quantitative and Qualitative Analysis)

Potential Harms Identified

Methodological Limitations

Active engagement in local green spaces

Gerber et al., 2017 [56]

United States; Bhutanese refuges recruited from local community garden and Bhutanese community network

Local community gardening at two local community plots. Authors clarified that these were in an urban area, and that some participants had to use the bus to access the gardens, but all were local to residents.

-
Gardening can boost connectivity and community strengths particularly for groups highly valuing communal functioning and cohesion
-
Expected outcome: Bhutanese gardeners would self-repot significantly fewer symptoms of distress and perceive higher levels of social support

-
50 people (62% female (n = 31))
-
mean age = 44.5, SD = 15.01
-
56% non-gardeners (n = 28), 44% gardeners (n = 22)

Quantitative data:
Structured questionnaires to collect cross-sectional data on:

-
symptoms of PTSD, anxiety, depression assessed by Refugee Health Screener-15 (RHS-15) in Nepali
-
Perceived social support using Medical Outcomes Study Social Support Survey (MOS SSS)
-
Somatization using the Patient Health Questionnaire-15 (PHQ-15), a 15-item self-report scale of somatization derived from the Patient Health Questionnaire, used previously with refugee populations, and translated into Nepali
-
Functional adaptation using Adapted Client Assessment Tool (ACAT) to describe current mastery or independence in 13 domains of functioning, orally administered

Qualitative data:
Semi-structured interviews to explore social support issues (such as the nature of social interactions whilst gardening), local acculturation (such as degree of adjustment to life in the US), and perceived advantages and disadvantages of the community garden.
Group meetings were held with participants to study findings and explore implications.

Descriptive analysis of quantitative data.
Adapted form of thematic analysis of interview transcripts for 8 gardeners and 4 non-gardeners, using the approach of Consensual Qualitative Research (CQR), which involved both deductive and inductive methods to code and interpret data.
Focus groups were then used to discuss findings of analyses of the data above to gain feedback on results and consider implications in a culturally salient way.

Quantitative:

-
RHS indicated that distress was not significantly different between gardeners and nongardeners
-
MOS SSS: gardeners reported greater social support (a moderate to large effect size (d = 0.70; 95% CI 0.12, 1.27), subscales revealed that gardeners reported significantly more tangible social support (large effect size, d = 0.88, 95% CI 0.28, 1.45), medium effect size was observed for emotional/informational social support (d = 0.53, 95% CI −0.04, 1.09), small to medium effect for affectionate support (d = 0.34, 95% CI −0.23, 0.89), and positive social interaction (d = 0.31, 95% CI −0.26, 0.86)
-
PHQ-15 scores found to be not statistically significant but found a small to medium effect size (d = 0.36, 95% CI −0.21, 0.91) with gardeners having more somatic complaints
-
Exploratory correlational analyses: Age was additionally significantly related to perceived emotional/informational social support (r = −0.38, p = 0.007; 95% CI −0.60, −0.11), with perceived support decreasing with age; ACAT was significantly related to time in US (r = 0.30, p = 0.033, 95% CI 0.02, 0.53)

Qualitative:
-
Key themes: General findings (results presented in all, or all but one), typical findings (more than half of the cases), variant findings (2 to 4 cases)

None

-
Self-selection to group membership leading to gardeners significantly more likely to live in a house and lower medical bills
-
Randomization of assignments may adversely disruptive to the ethos of the established community
-
Elderly with physical limitations may have difficulty obtaining certain benchmarks
-
Small to moderate effect sizes may be missed, although sample size was sufficient for hypothesis testing
-
Under-reporting of adjustment difficulties as participants feeling discomfort voicing their problems in interviews

Chiumento et al., 2018 [57]

United Kingdom; Children recruited from 3 schools in the North West of England
(two primary schools
and one secondary school)

Haven Green Space school garden project,
involving monthly sessions over the course of 6 months in which schoolchildren were supported at school by two horticulturists
and a Child and Adolescent Mental Health
Service (CAMHS) psychotherapist to work together in designing a green space

Promote positive mental, emotional and physical wellbeing of the children with the “Five Ways to Wellbeing” framework (Connecting with others; Being active; Taking notice of the local environment and of their feelings; learning horticultural skills and how to manage successes and failure; Giving back to the wider community

36 children (14 females)

-
children with behavioural, emotional and social difficulties
-
2 primary schools (year 5 and 6, aged 9–12) and 1 secondary school (year 7 to 9, aged 12–15)

Quantitative data:
Collection of pre- and post- intervention scores on the following measures for children

-
Mental wellbeing using Wellbeing check cards (based on the 7-item version of the Warwick Edinburgh Mental Wellbeing Sale)

Qualitative data:
collected over the course of 2 h workshops (pre- and post- intervention) by using the Mental Wellbeing Impact Assessment (MWIA) to plot data in a participatory way with children under the following three domains:

(1)
enhancing control
(2)
increasing resilience and community assets
(3)
participation and social inclusion.

This qualitative tool aims to assess the potential impact of a specific policy, service, project or program on the mental wellbeing of a population, and was originally developed for adults.

Quantitative analysis:
Statistical comparison of scores on pre- and post- intervention measures.
Qualitative analysis:
thematic analysis of data, with the coding process deductively driven by the MWIA themes.
Separate analyses of the quantitative and qualitative data then converged in the discussion to triangulate findings.

Quantitative:
-wellbeing scores not found to be statistically significantly different from pre- to post- intervention (although no test statistics were presented to support this statement)
Qualitative: Analysis of MWIA plots produced during workshops identified pre- and post-intervention tendencies towards pro-social behaviour (“feeling involved”, “having a valued role”, “sense of belonging” and “social networks and relationships”) and emotional symptoms. The thematic analysis also found that factors relating to mental health and wellbeing were positively impacted, including “emotional wellbeing” and “self-help”.

None

-
Lack of a valid control group and small sample size
-
Lack of validation of Wellbeing check cards in children or of the adaptation of the MWIA for children (in view of its
-
use of adult terminology)
-
Wellbeing check cards: limited in their intended ability to capture change over time

SD = standard deviation; CI = confidence interval.