Nature-based interventions for older adults: a systematic review of intervention types and methods, health effects and pathways

Abstract

Background

Nature-based interventions (NBIs) may support older adults’ health and well-being, but it remains unclear which interventions are most effective, for whom and in which contexts. The existing Wilkie and Davinson framework explains the pathways between NBIs and health outcomes but underemphasises the effects of environmental qualities. Therefore, the study aims to critically examine existing NBIs and their associated health outcomes, with a focus on identifying the environmental qualities and pathways that are either facilitating or impeding.

Methods

The review sought 16 databases for any nature-based interventions aimed at enhancing health outcomes where participants are older adults aged ≥65. The Mixed Methods Appraisal Tool assessed risk of bias. Narrative synthesis was used for result presentation.

Results

Of 6143 articles retrieved, 84 studies were included. Participants were primarily community dwelling without specific conditions. Most studies were quantitative experiments. Interventions were predominantly multicomponent, with restoring psychological capacities as the most common pathway. Spiritual, behavioural and socioecological changes were identified. Some health mechanisms were specific to outdoor interventions, but indoor interventions showed comparable outcomes.

Conclusions

The study predominantly included community-dwelling participants without specific health conditions, potentially limiting the generalisability of findings to older adults with multimorbidity or those in alternative living environments. Multicomponent interventions challenged direct associations between pathways and outcomes. Results extend existing frameworks by identifying spiritual, behavioural and socioecological benefits. Limited detailing of environmental qualities warrants further research to associate them with specific health outcomes.

Registration

The protocol was registered on PROSPERO (CRD42024496114).

Key Points

• The study identified nature-based interventions as having spiritual, behavioural, social and ecological benefits, additional to physical and well-being indicators.

• Limited research exists on small-scale environmental interventions.

• Indoor interventions showed broad benefits, whilst outdoor ones uniquely improved physiological health.

• Most participants were healthy community-dwelling older adults, limiting generalisability.

• Multiple activities in interventions complicated the understanding of specific mechanisms and outcomes, necessitating further research.

Introduction

The global population aged 65+ will double from 761 million in 2021 to 1.6 billion by 2050 [1]. This shift is most pronounced in Eastern and South-Eastern Asia, with low- and middle-income countries experiencing the greatest change, representing two-thirds of the world’s 60+ population by 2050 [1]. In the UK, 18.6% of the population was aged ≥65 in 2021 [2], with two-thirds of this group expected to live with multimorbidity by 2035 [3]. Understanding health factors, especially the physical environment’s role, is crucial for promoting healthy ageing [4]. Nature exposure has been shown to enhance physical [5–11], cognitive, and emotional [8, 12–14] and social health [15]. This evidence has sparked interest in nature-based interventions (NBIs) to improve public health [16, 17].

NBIs are ‘programmes, activities, or strategies engaging people in nature-based experiences for health and well-being’ [17]. Examples include prescribed outdoor walks and therapeutic horticulture. Key knowledge gaps remain regarding which NBIs work best for specific individuals and contexts [18], diverse intervention modalities (e.g. forest walking vs horticultural therapy) and environments (e.g. forest vs courtyard). Therefore, health professionals may struggle to determine the most effective approaches. This demands a better understanding of the pathways linking components of NBIs and specific health outcomes.

Whilst a recent systematic review provides an organised framework of pathways [19], older adults may experience these pathways differently, leading to varied health outcomes. For instance, neighbourhoods lacking activity opportunities could harm cognitive and physical health [20, 21], limiting their ability to build capacities such as exercises to enhance physical strength via NBIs. Besides, the Wilkie and Davinson framework [19] lacked consideration of physical environmental characteristics that can impact outcomes, with some studies showing better results in forests [19] than in parks [22].

Hence, the systematic review aims to address three key questions.

• What range and types of NBIs are available for older adults?

• Based on the framework by Wilkie and Davinson [19], what pathways link nature and health outcomes amongst older adults?

• What environmental qualities are associated with NBIs, and how do they facilitate or hinder pathways to health outcomes? [23]

Methods

This systematic review was conducted and guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 Statement [24]. The protocol was registered on PROSPERO (CRD42024496114).

Search strategies

The search strategy was developed in consultation with a university information specialist, drawing on relevant systematic reviews [19, 22, 25]. Searches were conducted across 17 databases, covering literature from 2003 to 1 February 2024. For details, see Appendix 1 in the Supplementary Data section for the full details of search strategies.

Selection criteria

Selection criteria followed the Population, Intervention, Comparison and Outcome (PICO) framework [26] (see Appendix 2 in the Supplementary Data section for the full details of selection criteria). Studies utilising virtual reality were excluded as these represent distinctive interventions involving immersive environments. However, virtual interventions may warrant a separate review for some populations such as older adults with limited mobility.

Title and abstract screening

Title and abstract screening was performed in Covidence [27] which automatically removed duplicates. Two authors independently reviewed titles and abstracts. Any disagreements were resolved by a third author.

Full-text review

Potentially eligible studies were retrieved as full texts, which were reviewed independently by two authors, and included if they met the criteria. Disagreements were resolved by a third author.

Data extraction and quality assessment

The Mixed Methods Appraisal Tool (MMAT) was used for quality assessment [28]. Data extraction covered study characteristics, intervention details, pathways, pathway domains, mechanisms, information on the physical environments and health outcomes. Only articles scoring >60% were included in the review (e.g. [29, 30]). For details, see Appendix 3 in the Supplementary Data section for the full details of data extraction and quality assessment.

Results

Study characteristics

The search identified 6143 articles, with 3264 duplicates removed, leaving 2879 for title and abstract screening. In total, 2636 were excluded at title and abstract stage, and 241 articles underwent full-text assessment; of these, 156 were excluded. Of the 87 studies appraised for quality, 3 were excluded for scoring ≤60% on the MMAT [31–33], resulting in 84 articles reported in the review. See Appendix 4 in the Supplementary Data section for PRISMA workflow. Table 1 summarises the characteristics of the studies included. Appendix 5 in the Supplementary Data section summarises the findings of the review. MMAT scores showed 39 studies scoring 100%, 24 between 80% and 90%, and 21 between 60% and 70%. See Appendix 6 in the Supplementary Data section for the full details of MMAT assessment scores.

Table 1.

Characteristics of included study

Study design	Count
Pre–post experiments with control conditions	32
Pre–post experiments without control conditions	23
Interviews or focus groups or questionnaires*	12
Randomised controlled trials	10
Participatory designs and evaluations	2
Case study	2
Quantitative survey	2
Observation*	2
*1 study utilised both observation and in-depth interview.
Research methods
Quantitative measurements	56
Mixed methods	15
Qualitative methods	13
Geographical location
Asia	50
USA and Canada	18
Europe	14
Oceania	2
Publication year
2000–2010	15
2011–2015	10
2016–2020	37
2021–2024	22
Participant residence
Community dwelling	46
Long-term care units, nursing homes, care homes or assisted living facilities*	23
Hospitals or medical health centres	9
Daycare centres*	5
Others: group homes	1
Not reported	1
*1 study has participants residing both in long-term care units and day care centres. Therefore, the overall number is >84.
Participant condition
Older adults without specific conditions	67
Older adults with dementia or Alzheimer’s disorder	9
Older adults with symptoms of depression	3
Older adults who were pre-frail or frail	2
Older adults with stroke	2
Older adults with hypertension	1

What range and types of NBIs are available for older adults?

The interventions were classified into three categories: (1) nature used directly [34–62], (2) nature in the background [63–88] and (3) both direct and background nature [89–115]. In background-only interventions, nature was mainly used to enhance the primary intervention [63, 64, 66–68, 70, 71, 74–78, 80, 81, 83, 85, 88]. Some used nature to promote healthy behaviours such as physical movement in green spaces [65, 69, 72, 79, 82, 84, 86, 87]. One study used natural scents to measure incremental changes in eye movement, heart rate and physical movement of participants with dementia [73].

Horticultural or gardening activities were featured in 50 studies [34–49, 51–62, 84, 87, 89, 90, 92, 94–99, 101–108, 112, 115, 116], typically encompassing plant life cycles like seeding, planting and harvesting (e.g. [42]). Whilst 19 interventions involved only horticultural activities [34, 37, 43, 45, 47–49, 54–56, 59, 94, 95, 97, 102, 104–106, 112], most interventions were hybrid [35, 36, 38–42, 44, 46, 51–53, 57, 58, 60–62, 84, 87, 89, 90, 92, 96, 98, 99, 101, 103, 107, 108, 115, 116], combining horticulture with educational elements [35, 38, 42, 46, 89, 92, 99, 107, 108] or arts and crafts [40–42, 52, 53, 57, 58, 60–62, 98, 103, 107, 108, 116], such as floral arrangements or making nature art (e.g. sketching leaves, vegetable printing).

Plant maintenance with advice and demonstrations on plant care was integrated into 11 studies [35, 36, 38, 42, 52, 60, 89, 101, 107, 108, 115], and 8 included social opportunities [87, 89, 90, 93, 96, 99, 107, 116]. Several interventions incorporated formal therapies like cognitive behavioural therapy [117], eye–hand coordination exercises [44] and reminiscence therapy [39, 83, 118]. Other hybrid activities involved animal care [89], food-making [40, 42, 46, 51, 52, 57, 87, 89, 98, 99, 107, 116], walking [64, 67, 78, 81, 98, 107, 108], gaming (such as nature-image sorting) [50], reflective writing [51] and walking [98, 107, 108].

For interventions using nature only in the background, walking (in forests, parks or greenspaces) was the most common intervention [63–65, 67, 70, 74, 75, 78, 81, 83, 88]. Walking was implemented as a stand-alone activity [65, 70, 75, 88], combined with mindfulness [63, 78] or integrated with other components, such as walking acupuncture (taping red beans to feet to stimulate acupuncture points [78]) or reflective writing [83]. Other activities featuring nature only in the background included physical exercise [66, 68, 71, 74, 76, 81, 83, 85], Taichi [76] and Qigong [64].

Some interventions involved upgrading the built environment [65, 79, 82, 84, 86, 87]. These upgrades included creating new planting areas, primarily raised planting beds, in existing outdoor spaces to facilitate horticultural or gardening activities [93]. Some focused on aesthetics by including new planting [72]. Some focused on accessibility by installing new wheelchair planters [82, 84, 93], rebuilding garden paths and surfaces [82, 84, 93], and installing a viewing platform overlooking the planting areas [84]. Some interventions aimed to improve comfort by constructing a conservatory for shelter and adding rattan blinds for shade [86]. One intervention focused on creating garden features to trigger memory and to engage with residents [84]. One focused on cost by installing low-cost decorations on trees in existing parks, hanging baskets of plants, planting new species to add visual and seasonal interest, and adding aromatic herbs to enhance sensory experiences [69].

The final main type of intervention involved actively and directly viewing nature from a fixed position. These interventions included observing a rooftop garden [73] or landscape areas in a park [79].

Based on the framework by Wilkie and Davinson [19], what pathways link nature and health outcomes amongst older adults?

Pathways, pathway domains and mechanisms

Based on the Wilkie and Davison’s framework (Fig. 1) [19], all included studies comprised some form of nature contact and/or nature experience. Most interventions utilised multiple pathway domains: reducing harm and restoring or building capacities [34–40, 42–48, 50–53, 56, 57, 60–78, 81, 83–103, 105–116, 119]. Most interventions restored psychological capacities [34–67, 69–79, 81–94, 96–104, 106–116, 118] by reducing stress or improving cognitive functions and/or emotions. Over half built capacities with physical activity and/or other health behaviours, primarily through increased activity (e.g. frequency of gardening) [35, 36, 38, 39, 44, 46, 48, 50–53, 56–58, 60–64, 66–71, 74–78, 81, 83–85, 87–89, 91, 92, 95–100, 105–116]. Many utilised social contact or interaction [34, 36–38, 40, 42, 43, 45–48, 50, 53, 57, 58, 64, 66, 69–72, 74, 76, 84, 86, 87, 89–94, 96, 99–103, 106, 107, 109, 113, 116]. A few aimed to reduce harm [64, 65, 73, 80, 91, 99] by immersion into an environment away from urban pollutants or enhancing exposure to health-promoting environmental factors such as sunlight and phytoncides. None reported harm caused.

Figure 1.

Pathways, pathway domains, mechanisms and public health outcomes of nature-based interventions. Adapted from Table 1 in Wilkie, S., & Davinson, N. (2021) [19].

All interventions have multiple mechanisms. Nearly all interventions involved cognitive restoration and/or inducing positive emotions [34–67, 69–72, 74–79, 81–94, 96–104, 106–116, 118]. Enabling environmental context or resources [35, 36, 38, 39, 42, 44, 46, 48, 50–53, 56–58, 60–64, 66–71, 74–78, 81, 83–85, 87–89, 91–93, 95–100, 105–116] for changes in behaviour or aid in performance for certain physical activities leading to specific health outcomes was the next most prevalent. Social influence to motivate behavioural performance was also common [34, 36–38, 40, 42, 43, 45–48, 50, 51, 53, 57, 58, 64, 66, 69–72, 74, 76, 84, 86, 87, 89–94, 96, 99–103, 106, 107, 109, 113, 116]. Some interventions involved education to develop skills [34–36, 38, 40, 42, 44, 46, 48, 50–53, 66, 76, 77, 89, 91, 92, 99, 101, 105–108, 116] and knowledge [34–36, 38, 40, 42, 44, 46, 48, 50–53, 66, 76, 77, 89, 91, 92, 99, 101, 105–108, 116]. Environmental factors like sunlight [64, 65, 67, 71, 91, 99] and phytoncides [64, 65, 67, 73, 80, 91] were utilised to promote health. One intervention targeted promoting confidence in horticultural tasks, involving beliefs about capabilities [35].

Public health outcomes

Most nature-based interventions promoted multiple public health outcomes [35, 37–45, 47, 49, 50, 53, 54, 56–60, 63–68, 70–77, 81, 82, 84, 85, 87, 89, 91, 93, 94, 96–99, 101–103, 105–107, 109, 110, 112–116, 118] with some promoting single health outcomes [34, 36, 48, 51, 52, 55, 61, 62, 78–80, 83, 86, 88, 92, 95, 100, 104, 108, 111]. Psychological well-being outcomes were the most common [34, 35, 37–45, 47, 49–65, 67, 68, 70, 73–77, 81–85, 89, 91, 93, 94, 96–100, 102–104, 106–110, 112–116, 118]. They included improving affect or mood, alleviating depression or anxiety, increasing health-related well-being, life satisfaction, attitudes towards ageing, perceived self-efficacy, self-esteem and reduced perceived isolation. Positive changes in physical outcomes included a reduction in cortisol [77, 80, 87], improved levels of heart rate [65, 68, 73], blood pressure [59, 65–67, 79, 82], chronic inflammatory indicators [67, 92], immune functions [80, 92], growth factors [95, 107, 120, 121] and scores in core physical functions (Barthel Index [38], Instrumental ADL [41, 58], Cup Stacking Test [44], Grooved Pegboard Test [57]), physical fitness (including aerobic fitness) [87, 96], balance ability [66, 87, 105], hand dexterity [96, 121], a lower chance of reporting falls and faster walking speeds [105]. Social outcomes included decreased loneliness or social isolation [35, 37, 47, 51, 99, 106, 122], increased social cohesion or social connection [47, 99, 106, 122], increased frequency of social contact [89, 122], social interactions [72, 98, 102, 109, 123] and social connectedness [37, 47, 99, 106, 107], a more positive sense of belonging [76, 122], improved interpersonal relationships [53, 93, 98, 114, 122], enhanced adaptation to separation [101] and larger social networks [122]. Some interventions showed enhancement in restoration with improved scores in cognitive functions, including Mini-Mental State Examinations [39, 42, 96, 118, 121], Clinical Dementia Rating [42], clock-drawing task [50], memory task [43], Working Memory Test [50] and Loewenstein Occupational Therapy Cognitive Assessment [39, 118].

Additions to the Wilkie and Davinson’s framework [19] were spiritual well-being, behavioural change for improved health outcomes and socioecological environment. Improved spiritual well-being was observed in both quantitative [93] and qualitative studies [47, 76, 84, 93, 106, 109, 112, 114]. Improved spiritual outcomes included enhanced insights or knowledge [75, 106] and better self-reports of spiritual well-being [93]. Reminiscing using nature was identified in four studies [84, 106, 112, 114]. Walking participants appreciated the beauty of nature, felt gratitude [75] and appreciated the environment more [106]. Taichi in a local park promoted a sense of control and optimism [76]. Finding a sense of purpose and meaning was noted in three articles [93, 112, 114]. Participants experienced higher anticipation and hope after a randomised controlled horticultural therapy programme [47]. Personal growth and development [47] and learning something new [106] were also noted. One indoor horticultural programme helped developed a higher sense of achievement [47]. Connection to something greater was identified in one study [112]. Finding peace was found in one Taichi park intervention [76] and another 3-month gardening project [112].

For behavioural change, agitation inventory scores decreased after introducing a therapeutic garden in a care home [84]. Horticultural therapy programmes increased verbal interactions [57] and participation in other gardening activities [113]. Decreased behavioural disturbances and improved acclimation to facilities were observed [49]. Enhanced engagement scores were observed in care home residents [50], who also tend to spend more time outdoors when gardens are improved [84]. Increased use of a conservatory garden in a hospital setting [86] and greater receptiveness to environmental changes [114] were also identified.

Positive socioecological outcomes were identified in one study concerning the construction of a therapeutic garden in a dementia care home [84]. The upgrade of the outdoor environment attracted 22% more visitors, and family members reported enjoyment when visiting the care home [84]. This increased visitation might encourage social interaction with residents and promote associated positive health benefits.

Activities, pathways, pathway domains and health outcomes

Figure 2 summarises interventions, pathways, mechanisms and health outcomes. Interventions utilising nature in both the background and the intervention covered the widest range of pathways and mechanisms. Direct nature contact interventions exhibited fewer pathways but broader health outcomes, excluding socioecological ones. Interventions using nature in the background, alongside other interventions, covered similar health outcomes as those utilising nature in both the background without other interventions, except for cognitive benefits. Nature introduced only in the background of interventions with interventions had the narrowest range of outcomes, with no evidence for functional, behavioural or socioecological outcomes.

Figure 2.

Summary of activities, pathways, pathway domains and health outcomes.

What environmental qualities are associated with NBIs, and how do they facilitate or hinder the pathways to health outcomes?

Figure 3 summarises the relationship between environments, pathway domains, mechanisms and health outcomes. Outdoor-only interventions utilised the widest range of pathway domains and mechanisms, whilst indoor interventions and mixed indoor–outdoor interventions were comparable in this regard. Despite fewer pathways, indoor interventions achieved similar health outcomes as outdoor interventions. Functional health outcomes were only evident in indoor interventions, whereas socioecological outcomes were evident only in outdoor interventions.

Figure 3.

Environmental types, pathway domains, mechanisms and health outcomes.

Linking specific environmental elements to health outcomes is challenging due to inconsistent reporting. For interventions used in the background [63–115], 31 lacked information on environmental characteristics [35, 45, 48, 49, 63, 64, 66, 70, 71, 74–78, 81, 85, 87–95, 98, 100, 101, 103, 105, 109, 111, 113, 115, 116]. Environmental characteristics reported varied in the level of detail. Some included only an overview of the environment [68] or the location only [97, 107, 108]. Some described the intervention and how it related to the environment [69, 80, 96, 99, 102, 104, 110, 112]. Some provided more details including species [65, 67, 72], layouts and features [73, 79, 82, 83, 114] and/or images [84, 86, 106].

Discussion

Range and types of nature-based interventions

This systematic review identified 84 empirical studies that integrated nature directly within the intervention, as the background or both. Horticultural activities emerged as the most prevalent intervention type. Psychological well-being was the most frequently examined and reported outcome. However, detailed descriptions of the environmental qualities associated with these interventions were notably limited.

For interventions with horticultural activities, these activities are typically not stand-alone but combined with other activities and arts and crafts are the most common hybrid activity. These hybrid activities resulted in consistent physical (e.g. increased muscle strength [124]) and psychological observations (e.g. reduced depression scores [38]). These findings suggest that interaction with nature can occur through direct planting activities and related crafts (e.g. making grass dolls and rocky leaf prints [40]), which are particularly relevant for settings such as care homes with limited outdoor space. Engaging older adults with nature through arts and crafts appears to deliver comparable health benefits to traditional horticultural or gardening activities.

For nature interventions not involving horticultural activities, hybrid activities were still more common than stand-alone interventions. Walking, often combined with mindfulness, exemplifies this trend. Walking as a physical activity consistently benefits physical and psychological health, with nature as a backdrop reducing perceived exhaustion [125] and potentially increasing activity duration and intensity. Some interventions combining walking and mindfulness specifically target older adults with depression [75, 100], indicating the adaptability of interventions to suit specific conditions. Mindfulness, as an evidence-based intervention for depression, suggests that it may be particularly beneficial for older adults with multimorbidity to avoid the adverse effects of polypharmacy [126].

Only one study [84] involved small design interventions to the physical environment, such as the installation of planters, whereas other studies required relocating participants to different settings or making substantial modifications to intervention areas. This underscores a research gap in assessing the impact of small-scale environmental modifications on health outcomes compared to more extensive changes, particularly in care settings such as care homes where space may be constrained.

Pathways linking nature and health outcomes

The study extended beyond the existing framework by Wilkie and Davinson [19], identifying additional benefits that their model did not encompass. These include spiritual, behavioural and socioecological dimensions. These benefits positively impacted health outcomes, highlighting the multifaceted advantages of NBIs. A summary of these additional outcomes is available in Fig. 4.

Figure 4.

Health outcomes additional to those identified in Wilkie and Davinson.

Approximately half of the interventions were controlled experiments. Besides assessing immediate physiological (e.g. reduced heart rate) and psychological (e.g. decreased depression) effects, many studies targeted health aspects important for healthy ageing. Notable improvements were observed in mobility and physical functions, such as the Barthel Index [38, 40–42, 82], visuospatial function [43] and balance [66]. Social outcomes, including reduced loneliness [36, 47, 51, 60], were also examined, alongside biological risk indicators for chronic conditions, like inflammatory cell counts [67], growth factors [95, 110] and immunity cell counts [116]. These findings suggest that NBIs can prevent chronic conditions and promote healthy ageing, supporting the recommendation of and the development of nature-based programmes in institutional settings for older adults.

Most interventions were multicomponent, making it challenging to untangle specific activities, their pathways and their effects. This complexity highlights the need for more precise studies to isolate the impact of individual components within NBIs.

Interventions that did not use nature in the physical environment still produced a range of benefits, indicating that positive health outcomes can be achieved through activities involving natural elements (e.g. pot gardening, flower arranging, arts and crafts). This is especially relevant for older adults with limited mobility, for whom accessing outdoor spaces may be challenging.

However, interventions that used nature in the physical environment provided a greater range of benefits than those where nature was merely integrated into the intervention. In particular, these interventions reduced harm by mitigating air pollution, abating heat and noise, and providing beneficial microbiota, phytoncides and sunlight—effects not observed in interventions where nature was integrated and not in the physical environment.

One NBI using nature in the background could yield socioecological benefits, positively affecting stakeholders and the overall environment, thereby providing additional indirect benefits to the target population. For example, the construction of a therapeutic garden in a dementia care home increased visits by 22%, with family members reporting greater enjoyment when visiting the care home [84], therefore showing the potential for exploring indirect positive impacts on residents by enhancing benefits on people within the socioecological environment.

Environmental qualities and their roles

Functional health outcomes were only observed in indoor interventions, suggesting that certain environments may favour specific benefits or that other functional outcomes lack measurement. The lack of details about the physical environment makes it challenging to link environmental characteristics and specific health outcomes.

Limitations and future directions

The review extends Wilkie and Davinson’s framework [19] by identifying other health outcomes important for healthy ageing. Its strengths include the comprehensive search strategy, independent assessment and pre-registered protocol. Limitations include the complexity of interventions with multiple activities, making interactions hard to analyse. Moreover, most studies focused on community-dwelling adults without specific conditions and non–community-dwelling environments, limiting insights for older adults with multimorbidity and other physical environments. Additionally, few studies detailed environmental qualities, necessitating further research to connect these with health outcomes.

Conclusion

The study extended beyond the existing framework by Wilkie and Davinson [19], identifying spiritual, behavioural and socioecological health outcomes. Whilst indoor interventions offer diverse benefits, the environment of outdoor NBIs improves physiological health. Evidence supports the benefits of NBIs, though most studies involved community-based participants without specific conditions. Health outcomes may vary, especially for those with multimorbidity or living in other housing or care environments. Interventions often included multiple activities, making it difficult to distinguish specific mechanisms and outcomes, warranting further research.

Declaration of Conflicts of Interest:

None declared.

Declaration of Sources of Funding:

This systematic review is supported by funding from the Advanced Care Research Centre, University of Edinburgh. J.L. is funded by the National Institute for Health and Care Research (NIHR) Applied Research Collaboration (ARC) North East and North Cumbria (NENC) (NIHR200173). The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care.