Can Physical Activity Support Grief Outcomes in Individuals Who Have Been Bereaved? A Systematic Review

Background In 2018, there were 616,014 registered deaths in the United Kingdom (UK). Grief is a natural consequence. Many mental health concerns, which can be identified as grief outcomes (e.g. anxiety and depression) in those who have experienced a bereavement, can be improved through physical activity. The objective of this review was to identify from the existing literature if physical activity can benefit grief outcomes in individuals who have been bereaved. Methods A systematic review of nine databases was performed. Included studies (qualitative and quantitative) explored physical activity to help individuals (of any age) who had experienced a human bereavement (excluding national loss). Results From 1299 studies screened, 25 met the inclusion criteria, detailing eight types of bereavement (parental (n = 5), spousal (n = 6), patient (n = 4), pre-natal (n = 3), later life (n = 1), caregiver (n = 1), multiple (n = 4) and non-defined (n = 1). Activities including yoga, running, walking and martial arts were noted as beneficial. Physical activity allowed a sense of freedom, to express emotions, provided a distraction and an escape from grief, whilst enhancing social support. Conclusion There is some evidence that physical activity may provide benefit for the physical health and psychological wellbeing of those who have been bereaved, including when the loss has happened at a young age. This review is timely, given the wide-scale national loss of life due to COVID-19 and extends knowledge in this area. More research is needed to explore the benefits of physical activity for those who have been bereaved. In particular, there is a need for well-designed interventions which are tailored to specific activities, populations and grief outcomes.


Background
There were 616,014 registered deaths during 2018 in the UK alone [1][2][3]. Bereavement is a common experience following death [4] and is a term often used interchangeably with grief and mourning; yet they differ in meaning. Following a bereavement, individuals grieve and mourn in different ways. Bereavement refers to being in a state of loss as the result of the death of a significant relationship, mourning is an expression of grief and grief is the complex bio-psycho-social response and reaction to bereavement [5,6]. Individuals may experience a range of grief outcomes following a bereavement such as increased levels of anxiety and depression [5], self-harm [6], alcohol and drug use [7] and suicide ideation or attempts [8]. Individuals may experience a decrease in concentration [9] and self-esteem [10], or they may experience insomnia, aggression [11], or post-traumatic stress [5]. Individuals can experience multiple grief outcomes simultaneously, at different rates, intensities and durations [12]. There are a multitude of factors which can influence how a person grieves; these include their age, the type of death and their relationship to the deceased [5,13]. Children and young people have a different understanding of death to adults, and they may not fully understand the situation or their feelings relating to it [5,9,12]. Experiencing a bereavement as a result of murder may leave individuals with increased aggression, guilt, unanswered questions and traumatic imagery [14,15]. This type of death can additionally delay the grieving process as the family may not be able to grieve during the criminal justice trial [16]. Some bereavements, for example those from death by suicide, may also lead to prolonged grief disorder (PGD), a term used to describe an ongoing, heightened sense of mourning and rumination [17], leaving individuals with an increased sense of guilt [18]. PGD is categorised as grief outcomes which fail to diminish after 6 months causing disruption to daily, social and occupational functioning [17,19]. Experiencing a traumatic bereavement such as death by suicide as a child or young person can increase suicidal ideation and attempts when compared to those who experienced bereavement due to a natural death [20,21].
Those experiencing grief outcomes from both natural and sudden or traumatic bereavements should be able to seek appropriate support. Within the UK, there are nationwide and local regional bereavement services catering to all ages. Services such as Cruse, Child Bereavement UK, Hope Again and Winston's Wish have centres around the UK, with CHUMS, Simon Says, Halo and several more offering support in specific localities. Each of these services offer a variety of bereavement support, including individual or group counselling, telephone support, activity days and weekend retreats.
As many grief outcomes such as depression, anxiety, anger, lowered self-esteem, substance use, self-harm and suicide ideation are also mental health concerns, without the experience of a bereavement, it is plausible to suggest that options to improve mental health may well improve these issues when manifested as grief outcomes. One way to improve mental health is through physical activity; and a number of studies have found that it can benefit and reduce factors such as depression [22][23][24] and anxiety [25][26][27]. Physical activity has been shown to reduce aggression [28,29], improve life satisfaction [30] and reduce post-traumatic stress disorder (PTSD) [31]. Adventurous physical activity (e.g. rock climbing) improves positive and negative affect and self-efficacy [32]. However, the extent to which physical activity may be beneficial to such outcomes, in those who have been bereaved, is unknown.
This review aims to identify studies which have investigated the role of physical activity following a bereavement, with a specific question as to whether physical activity can benefit grief outcomes. It further seeks to understand what grief outcomes are commonly recorded, how they are measured and the types of physical activities reported to be helpful to those who have experienced a bereavement.

Protocol Registration
This review is reported using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A protocol for this review has been registered with PROSPERO (Ref: CRD42017081237).

Eligibility Criteria
Eligible studies were not restricted by year, nor by study design, but were restricted to those written in the English language. Study characteristics were defined using PICOS (Population, Intervention, Comparator, Outcome, Study design [33]) as below to identify relevant sources.

P-Population
Population were of any age and sex and must have experienced personal grief (e.g. death of someone they knew). Studies that described those who had experienced national grief (e.g. '9/11' or death of a monarch) were excluded. Populations grieving the loss of a pet, activity or a sporting injury (e.g. loss due to retiring from sports) were also excluded.

I-Intervention
Studies were included if physical activity was used to support bereavement, grief or mourning, or if physical activity was included as a measurement in relation to bereavement. Physical activity was defined as any activity that uses the skeletal muscles and requires an energy expenditure of above 1.5 metabolic equivalent (MET), which is considered resting [34]. Physical activity could range from light to vigorous intensity and be of any duration or mode (e.g. walking, football, dance). Studies which used physical activity with the primary aim to prolong life, reduce premature death, or linked to mortality were excluded.

C-Comparison
This review did not restrict inclusion by comparator.

O-Outcomes
The primary outcomes were cognitive, affective and behavioural grief outcomes. Cognitive/affective grief outcomes included anxiety, depression, anger, guilt, loneliness, post-traumatic stress, self-esteem, suicidal ideation, well-being, resilience, life satisfaction and quality of life. Behavioural grief outcomes included alcohol or drug use, emotional eating, isolation, insomnia, selfharm and suicide attempts. Other outcomes not listed but that were attributed as grief outcomes were documented.

S-Study Design
All study designs were included in this review.

Information Sources
This systematic review was performed during May 2019 and updated in November 2020 and includes studies from the inception of the databases. Initially, the Cochrane Library and PROSPERO were searched for relevant reviews to avoid duplication. A comprehensive search of the following electronic databases was then performed: BASE, Directory of Open Access Journals, Medline PubMed, PsycArticles, PsycINFO, Science Direct, Scopus and SPORTDiscus. Once all appropriate articles were identified, cited references were hand searched for further appropriate articles.

Search Strategy
A combination of terms from medical subject headings (MeSH) and keyword variants were used to identify records. Search terms found in Table 1 were used and adapted to each database search.

Study selection
Records were imported into Mendeley [35] software with any duplicates removed by JW. All titles and abstracts were screened for clear violation of inclusion criteria followed by 100% full text screening for eligibility independently by JW and GWS. Any discrepancies in decision making were discussed and resolved with AMC.

Data Collection Process
Two reviewers (JW, AMC) independently extracted and inserted data from 100% of the articles into Excel, using a predefined data extraction form. This form was piloted and updated to include additional extraction categories for clarity using two initial studies. The final version included study title, authors, journal, publication date, country of study, study design characteristics (including type of study and any comparator groups), recruitment method and funding source. Participant characteristics included sample size, demographics (age, sex, ethnicity, religion), relationship to deceased, length of and mean time since the bereavement. Interventions were described using the TIDieR (Template for Intervention Description and Replication) checklist and guidance framework [36], where relevant information such as intervention fidelity, tailoring, adaptions, provider, number of sessions, procedure, delivery, materials, type and length of the intervention were recorded. The type of physical activity, measures used, theoretical frameworks, follow-up periods and any behaviour change techniques [37] were recorded. It was noted if a control group was used. Outcome data recorded the aim, main outcomes, grief outcomes and measurements used.

Risk of Bias in Individual Studies
Two reviewers (JW, GWS) independently assessed the quality of the eligible studies using the Mixed Methods Appraisal Tool (MMAT) [38]. The MMAT supports the assessment of the quality of qualitative, quantitative (randomised, non-randomised and descriptive) and mixed methods studies.

Synthesis of Results
The impact of physical activity with any available measurement of the primary grief outcomes was reported. A meta-analysis was not possible due to the variety of outcomes and study designs; therefore, a narrative synthesis was performed.

Study Selection
A total of 1545 titles and abstracts were initially identified from the search criteria, with an additional two included from hand searching. Of these, 248 duplicates were removed, leaving 1299 records to be screened. Of these, 1248 titles and abstracts were excluded, leaving 51 articles for full text review. A total of 25 of these articles met the inclusion criteria and were included in the analysis (see Fig. 1).

Participant Characteristics
A total of 157,068 individuals participated across the 25 articles, with an age range of 6-91 years old. Studies included 47.5% males and 52% females; 0.5% could not be attributed as two papers failed to provide details on participant sex. The high overall sample size is due to two studies with a total of 139,097 [62] and 12,759 [51] participants respectively. Four ethnicities were reported across 12 studies, specifically Hispanic, Caucasian, Asian and African-American. Thirteen studies did not mention ethnicity [42, 43, 45-47, 52, 55-62] or just stated 'other' [48].

Length of Bereavement
The time since death related to the bereavement varied between studies, from 0 months to 15 years. Length of time since death was omitted from six studies [41,46,49,51,52,55,59,61]. The longest length of time since death was 15 years [53,54]. Eleven studies examined recent bereavement up to two years since death. Length of time participants had known the person who died was not routinely reported.

Types of Physical Activity
There were 22 different physical activities mentioned. Brewer and Sparkes [53,54], through interviews, found activities such as martial arts, rugby, football, running and walking were beneficial following parental bereavement. Others [44] found canoeing, hiking, treasure hunts, and other play activities helped with PTSD in young people who had been bereaved of a parent. These findings were supported [49], with traditional camping activities such as canoeing being found to be helpful after parental bereavement. Dan-Jeon breathing and stretching was found to decrease levels of stress and increase life satisfaction [42,45]. Qigong (similar to Tai Chi) was found to reduce feelings of fatigue [43]. Physical activities such as yoga, walking and running were reported [40,47,48] to be useful in supporting grief following pre-natal death (still-birth). Walking was also found to be supportive for widowed individuals following the death of a spouse [60]. The above studies provided details of the type of physical activity used or mentioned within the research as it was a main aim or focus. Others without such focus found activities including sports, spending time outdoors, cycling, hiking, yoga and running were effective coping strategies used by medical practitioners after the death of a patient [56][57][58]. A number of studies failed to record type of physical activity when investigating its impact on grief outcomes. These were often studies which used cross-sectional designs, accessed large databases or used health behaviour questionnaires that were asking about generic physical activity performance [39, 41, 46, 47, 50-52, 55, 56, 58, 59, 61, 63].

Behavioural Change Techniques Used in Interventions
Commonly, eligible papers did not provide enough detail on techniques related to increasing physical activity. Two studies [42,45] report on the same project whereby intervention participants were given a demonstration led by a Dan-Jeon master, weekly instructions on performance, and the opportunity to practice stretching activities. Using the behaviour change technique (BCT) Taxonomy Version 1 [64], these studies used BCTs: 6.1 (Demonstration of the behaviour), 9.1 (Credible source), 4.1 (Instruction on how to perform the behaviour) and       8.1 (Behavioural practice/ rehearsal). Another study [39] used the same four BCTs, with the additional BCT 2.3 (Self-monitoring of behaviour). Within this study, participants were provided with a demonstration, instructed for the first 5 weeks, practiced stretching exercises which were led by a Qigong master and self-monitored their physical activity. One study [39] focused on digital monitoring of health behaviours, using BCTs 2.2. (Feedback on behaviour) and 2.1 (Monitoring of behaviour without feedback). Another study of an online yoga programme [40] used BCT 9.1 (Credible source), utilising a reliable source for yoga videos. BCT 12.5 (Adding objects to the environment) and 4.4 (Instructions on how to perform behaviour) were also used within the intervention, by providing yoga equipment, and instructions on how to use the videos. BCTs were unable to be coded for all other studies, often as they were not interventional in nature.

Summary of Studies
A synthesised narrative of the overall results provides evidence which suggests that physical activity can support grief outcomes following a bereavement. Physical activity was found to allow those who have experienced a bereavement to express emotions, escape from grief, retain memories and gain a sense of freedom [53,54]. Participating in physical activity allowed individuals to create friendships, drawing upon social support, whilst also creating closer family cohesion [49,54]. Physical activity enabled individuals who have experienced bereavement to reduce levels of depression, stress, loneliness and PTSD [42,44,48,59,63]. Medical professionals used sport and physical activity as a coping strategy after experiencing patient death [52,[56][57][58]. Engagement in physical activity was, however, reported to reduce following a bereavement [60], with some identifying barriers such as guilt, lack of time, fatigue and no motivation to be active, for this reduction [47]. However, it was acknowledged that physical activity levels increased with time [60] and that once barriers were overcome, it helped individuals to feel better following a bereavement [48].

Discussion
This is the first review of its kind that suggests that physical activity may be beneficial for individuals following a bereavement, with some evidence that it can benefit several grief outcomes. This was found across different types of bereavement including parental, spousal, foetal and patient. Whilst there was general agreement between the studies that physical activity provided some benefit following bereavement, this benefit could not be matched to one single grief outcome. Depression was the most commonly mentioned grief outcome, and others most often fell within the 'mental health' umbrella (e.g. anxiety, stress, guilt, negative affect). This supports previous research outside of the bereaved populations, which shows physical activity to be beneficial to mental health [26,29,[65][66][67]. Some common grief outcomes were not measured in these studies, for example alcohol consumption or self-harm. As such, the focus on grief outcomes is quite narrow, and important impacts of bereavement are understudied. This review calls for more high-quality research in the area of physical activity and bereavement, particularly given the current surge of international bereavement since the start of the COVID-19 pandemic. The type of physical activity most likely to benefit grief outcomes could not be easily determined, as there were so many types presented in the research. However, they can be summarised under those linked with the outdoors (e.g. walking, running, hiking, canoeing, cycling), activities around relaxation and a focused mind (e.g. yoga, Dan-Jeon, Qigong, mixed martial arts) or team sports (e.g. football, rugby). Most of these were evaluated using qualitative methods, and there was a lack of quality randomised controlled trials evaluating these activities as interventions compared to a comparator group. These studies show that physical activity to support bereavement is promising, yet more rigorous, experimental research is needed. Many studies failed to define type of physical activity, which can impact replication for future interventions. Future work should consider which type of physical activity is most beneficial, whether there are differences between individual as compared to team or group activity, how physical activity impacts upon grief outcomes and whether this differs depending on type of bereavement.
Due to heterogeneity in study design, types of death encountered, physical activities performed and grief outcomes reported, the synthesis of the data extracted from this review was limited to a narrative approach. Many of the studies had low sample sizes, and three studies used the same cohorts of participants or intervention [42,44,45,49,53,54]. Considering that 616,014 deaths were registered within the UK in 2018 [1][2][3], it is important that future research draws from larger populations with better quality research methods, to provide a wider generalisation of results. Only a small number of studies (n = 7) that met the inclusion criteria for this review used an experimental/intervention design; therefore evidence from studies with an appropriate control group is limited, making it difficult to draw concrete conclusions. We know the impact of grief is substantial. For example, scholars have noted physical health changes following bereavement such as higher likelihood of hospital visits and longer stays in hospital [68], or impacts on productivity and functioning at work [69]. The need for appropriate interventions is increasing in importance as populations are living longer and we know the age at which a bereavement is experienced can impact the understanding of death and experience of grief outcomes [5,9,70]. The review examined participants from aged 6 to 91 years old; however, no paper distinguished the differences between age groups and their different reactions or understandings about death. Those experiencing bereavement in older age may perhaps have other stressors to contend with at the time of bereavement and their needs may be different to those of a younger age [71]. The types of physical activity identified in the review are inconsistent, with both team sports and individual sports identified; therefore, no affirmative conclusions could be drawn about the type of physical activity and the support it may provide to grief outcomes. Additionally, a number of studies failed to provide details on physical activity in terms of what activity was performed, the duration, intensity and frequency, thus limiting the conclusions we can reach in this review. We present the aggregate of all ages in this review, but the difference between grief outcomes and abilities or suitability of different types of physical activity at different ages should be explored in future work.
This review also attempted to describe physical activity interventions that aim to support bereavement, using BCTs. However, only five studies were able to be coded. BCT coding allows for better replication of an intervention or programme of change, and future researchers should make every effort to specify the content of their designs and/or findings using such reporting practices. Conclusions drawn from the use of the MMAT quality assessment tool should also be taken with caution. Whilst half of the studies were deemed to reach 100% in relation to quality as measured by this tool, there were clearly issues in relation to small sample sizes, study design, and potential replicability of findings.

Conclusion
This review was the first to systematically investigate the role physical activity plays in the lives of those who have been bereaved. By allowing broad search terms, this review was able to evaluate all types of bereavement, with any age, sex, relationship, and grief outcome. Yet even with such a wide lens, only 25 studies met the inclusion criteria. Our review of these studies suggests that physical activity may be one approach to consider for future intervention, with some evidence of its ability to provide benefit to individuals who have experienced bereavement, often having a positive impact on grief outcomes related to mental health. This often occurs in outdoor activities, those with a team nature or those that enable mindfulness and relaxation. However, physical activity levels have been found to decrease directly following a bereavement when there is no structured physical activity or intervention available. To improve grief outcomes and the impact of bereavement, more research is needed into which physical activity interventions are currently available for individuals who have experienced bereavement, how physical activity can support those who have been bereaved, which type of physical activity is used most by the population, which grief outcomes are most improved by physical activity and whether this differs by age, sex, type of death, relationship to the deceased, length of time since death and type of physical activity. Given the recent scale of international death due to the COVID-19 pandemic, there is an urgent need for wider research is in this area. In conclusion, evidence here suggests that physical activity may be a beneficial behaviour for those who have experienced a bereavement and should be considered as a priority for future research and for future interventions.