Mood andwell-being of novice openwater swimmers and controls during an introductory outdoor swimming programme: A feasibility study

1 School of Sport, Health and Exercise Science, University of Portsmouth, Portsmouth, UK 2Medical Statistics & Epidemiology, School of Health and Care Professions, University of Portsmouth, Portsmouth, UK 3 Radius Healthcare, Hove, UK 4 Anaesthetics BSUH, Royal Sussex County Hospital, Brighton, UK 5 Brighton and SussexMedical School, University of Sussex, Brighton, UK 6West Recovery Team, Sussex Partnership NHS Foundation Trust, Hove, UK

by the final session. Tension scores peaked in both swimmers and controls immediately before the first outdoor swim. Nonetheless the swimmers' improvement in mood and well-being scores was significantly greater than that of the controls. The nature of the study does not provide mechanistic understanding; there are likely to be a number of explanations (physiological, psychological and sociological) for the changes in mood and well-being in swimmers and controls that can be investigated further.

INTRODUCTION
Current physical activity for health guidance states that each person should take at least 150 min of moderate to vigorous physical activity each week. 1 However, globally, 23.4% of males and 31.7% of females would be deemed physically inactive. 2 The benefits of physical activity are well documented and include a reduced risk of developing both physical and mental health problems and can support treatment of preexisting health conditions. 3,4 Therefore, it is important that sedentary individuals find a form of activity that they enjoy, are motivated to continue and thus more likely to adhere to, 5 consequently, enhancing quality of life and reducing the healthcare burden.
Land-based activities such as Park Run have introduced and encouraged large numbers of non-runners or occasional runners, particularly from sectors of the population that traditionally have lower activity levels, for example, women, overweight people and older adults. 6 The regular use of outdoor environments for exercise or physical activity provides opportunities for changing indices of mood, well-being and mental health. [7][8][9][10] In addition, Mitchell 9 suggests that exercise in natural environments is associated with a greater reduction in the risk of poor mental health in comparison to other built environments. Therefore, it appears that there is a role for 'green' activity or exercise in improving and maintaining good mental health, and reaches populations that may otherwise be inactive.
Similarly, a large body of anecdotal evidence exits suggesting a link between outdoor swimming and improved mental health. 11 Many anecdotes discuss the 'post swim high' swimmers feel after their swim. 12 In fact this anecdotal information may be part of the reason for the growth in popularity of the activity as both a sport and pastime. [13][14][15] Qualitative enquiries present evidence of a therapeutic influence, in that the lived experiences of outdoor swimming elevated both physical and mental health and well-being 10,[16][17][18] and were transformative, connecting and re-orientating for participants. 19 Previous research has surveyed and compared cold water 'winter' swimmers and age-matched controls before and after the winter swimming season. 20 They found no differences between the two groups in mood before the winter swimming season, but found significant reductions in fatigue and increased vigour in the swimmers after the winter swimming season. The use of independent groups means that it is not clear if the changes in mood observed occur as a consequence of winter swimming or other differences in the groups, and if the changes in mood occur acutely or take time to manifest. Therefore, it was hypothesised that there would be acute as well as longer term changes in mood and well-being in a group of novice outdoor swimmers as a consequence of participation in an introductory outdoor swimming programme.
sea swims and fully participated in all sessions. The duration of stay in the sea was self-determined and subject to the weather and sea conditions. The coaching team ensured the health and safety of all swimmers during and after their swimming sessions. During each session, the controls either sat or stood watching the swimmers. For pool-based sessions, controls were away from the pool side on the swimming pool balcony; and during outdoor sessions they were on the beach, close to the swimmers clothing or bag.

Outcome measures and procedures
The abbreviated Profile of Mood States (POMS) questionnaire 21 and Short Warwick-Edinburgh Mental Well-being Scale (SWEMWBS) 22 were administered to swimmers and controls before the first pool session, the first sea swim (session 4), and the final sea swim (session 10). The POMS was also administered immediately after the same sessions to both swimmers and controls. The controls were inactive during the sessions and either stood or sat for the duration on the swimming pool balcony (session 1) or on the beach (sessions 4 and 10).
The abbreviated POMS is a 40-item questionnaire that asked participants to rate their mood across seven subscales (tension, fatigue, depression, anger, confusion, esteem and vigour) on a 5-point scale ranging from "not at all" (0) to "extremely" (4 These comments were grouped into themes.

Statistical analysis
The data were analysed using SPSS v25 and checked for normality of distribution; non-normally distributed data were compared using

Attrition
Of the 64 swimming participants who entered the study (Figure 1), three withdrew from the programme after the pool swimming session and the remainder of missing data occurred due to non-attendance at sessions in which surveys were conducted. The complete data set consisted of 49 participants who returned questionnaires at each stage.
This represented a 76.5% retention rate. Prior to the start of the study, all volunteers had no experience or exposure to swimming in cold sea water (water temperatures ranged from 15 • C to 21 • C). Twenty-two of the 24 non-swimming controls were retained (91%).

Chronic response (between session)
Between sessions comparisons for POMS and SWEMWBS subscale scores are shown in Tables 2 and 3 In swimmers, adjusted total SWEMWBS scores were significantly increased between the pool and final sea swim (P < .001, r = .75) and the first and final sea swim (P < .001, r = .53; Table 3). Whereas in the control group, increases in adjusted SWEMWBS scores were found between the first and final sea swim (P = .019, r = .43); there was also a trend toward a reduction in adjusted SWEMWBS total scores between the pool and first sea swim (P = .053, d = 0.34). Finally, significant associations in well-being level change were found between the pool swim and the first sea swim (χ 2 (3) = 13.148, P = .004, V = 0.393) and between the pool and final sea swim (χ 2 (3) = 54.791, P < .001, V = 0.803), showing increased well-being level in greater numbers of swimmers than would be expected, but not controls. Table 4 shows POMS score comparisons between swimmers and controls. Subscale scores before the swims did not differ between swimmers and controls; however, before the final sea swim, the TMD was significantly lower in the swimmers compared to controls (P < .001,

Swimmers verses controls (between group)
In addition, following all swim sessions, significantly higher Vigour and Esteem scores were found in swimmers compared to controls (all P < .001, d = 0.9-1.8 or r = .4), and Tension, Anger, Depression, Confusion and TMD scores were significantly lower in swimmers compared to controls (all P < .001 to P = .023, d = 0.6-3.8 or r = .2-.6).
The adjusted total SWEMWBS scores ( Table 5) indicate controls had significantly higher scores before the pool swim (P = .046, r = .21), whereas by the final sea swim, the swimmers had significantly higher adjusted total SWEMWBS scores than controls (P = .034, r = .23). At the final sea swim, the relationship between group (swimmers and controls) and the well-being level was significant (χ 2 (2) = 7.428, P = .029); swimmers were more likely than controls to have moderate well-being values ( Table 5). No other significant associations were found during the pool swim or the first sea swim.  a, pool versus first sea swim; b, pool versus final sea swim; c, first versus final sea swim. *P < .05; y = difference between swimmers and controls P = .004; z = difference between swimmers and controls P < .001.

Narrative
Upon completion of the course, volunteers were free to comment on their experiences, stating they found sea swimming challenging to begin 'it was tough to start with' , 'weirdly claustrophobic to begin with' or 'tiring and hard work' and all stating they would continue to sea swim following the course as they found the swimming 'invigorating' , 'enjoyable' ,

DISCUSSION
This study tracked the changes in mood and well-being during a    Therefore, these data reflect a broad spectrum of the population in terms of well-being and possibly mental health. The well-being scores of controls decreased between the pool and the first sea swim and increased again by the final sea swim, maintaining well-being from start to finish. In contrast, swimmers well-being scores increased at each stage, especially between the sea swimming sessions. However, this study can offer no mechanistic insight and these changes may have occurred due to a range of possibilities (physiological, psychological, and sociological) either as single factors or more likely a combination of factors.
The responses of the controls are also of interest. This group includes friends or family of the swimmers, observing their activities from the viewing gallery at the pool and from the beaches during sea swims. Compared to the pool session, the controls overall TMD was similar and total SWEMWBS score increased. However, these improvements in mood and well-being were not as large as those observed in the swimmers. The fact the controls were on the beach may have reduced negative mood states. There is evidence that proximity to water, whether you are entering the water or not, or visiting 'blue space' contributes to improved mental health. 28,29 In addition, before the first sea swim, the controls also experienced the same increase in tension as the swimmers. The controls were supporting the swimmers in a challenging activity, which may not initially appear to be that rewarding, but ultimately it was for the swimmers. The tension initially felt by swimmers may have been transmitted to their friends/family (acting as controls) whose tension levels appear elevated in comparison to the pool swimming session. Reflecting on the narrative comments, it may be that controls experienced vicarious emotion or empathy and projected their thoughts and feelings on to the swimmers positions. 30 The narrative accounts recorded after the course indicated the swimmers enjoyed their outdoor swimming experience. These accounts agree with previous observations that outdoor swimming offers a change in personal focus and has a calming influence, 11,19 and all would continue to participate in outdoor swimming. This is encouraging as the problem with many interventions is that participation rapidly declines over time. Although the study provides no mechanisms for the changes in mood or well-being, swimmers explained the positive effects of performing an exercise they came to enjoy. This included being in cold water, in an outdoor space with few people around, provided an opportunity for mastery, goal achievement, confidence building, challenge and a sociable environment with like-minded people. These explanations have been suggested previously for other forms of leisure activity 31 or exercise taken outdoors 7,9,17,18 as well as outdoor swimming. 11 It may also be that a number of these factors are interlinked. Previous research indicates that physical activity in natural environments is associated with a reduced risk of mental ill health, 9 and further to that, may have a positive therapeutic effect. 17,18 Considering a recent case study, 11 it seems pertinent to establish if outdoor swimming has any benefits to support improvements to mental health. Therefore, it would be of interest to determine how outdoor swimming and continued participation may affect people living with a clinically diagnosed mood disorder or depression.
Indoor and outdoor swimming are not alone in the acute and chronic effects on mood and well-being. Many forms of exercise appear to result in similar effects including treadmill running 32 and running outdoors. 33 Therefore, the potential benefits of outdoor swimming for improved mental health do warrant further enquiry. This is an activity that involves risk. Therefore, future studies should look to include terrestrial activities, as control groups, which may be more easily achieved and conducted more safely to establish if similar effects can be found.
These control groups could include cohorts that are active and sedentary in outdoor built and natural settings, and possibly cohorts who are active and sedentary indoors. By including these cohorts, we may start to tease out the separate effects of mode of activity or inactivity and the environment and which combinations are most and least effective in improving mood and well-being. However, this study does seem to add to the evidence that there is no one size fits all approach to improving mood, well-being and mental health. 34  The study is not without limitation. The first pool swim and first sea swim were 3 weeks apart and formed the start of the course; therefore, new group dynamics, unfamiliarity and participation in a new and challenging activity may have impacted on well-being and were evident in the elevated POMS Tension scores of swimmers and controls before the first sea swim. The comparison between the first and final sea swims was 6 weeks apart and formed the later longer part of the course. Therefore, the short time frame (3 weeks) between the first two assessments (first pool swim and first sea swim) may be too short for changes in well-being to be observed. Further to this, the Cronbach alpha value for the POMS scale suggested low internal consistency for the depression subscale (0.664). Despite the lower internal consistency, significant reductions were found in the depression subscale.
The shorter POMS form was used in this instance to maximize response rate as time at the start and end of each session was short. Using longer versions with greater internal consistency may have reduced the problem, but may have reduced the response rate, due to the increased completion time. In addition, although it would be preferable to recruit an equal number of male and female volunteers, this was not possible as the majority of the customers attending the courses were female.
Gender comparison studies have been performed with mixed results, some indicating no or minimal differences in mood state between males and females, 35,36 and others have found some indices to be elevated in females. 37 Furthermore, the nature of the convenience sampling method is a weakness as there was no randomisation to the study, but is a starting point from which stronger randomised control trials (RCTs) can build upon. In the present study, swimmers were paying for 10 coached sessions in an activity they expressed an interest in doing, thus they were motivated to attend the sessions. However, retention in RCT may be lower than that in the present survey, due to the participants being allocated randomly to either the intervention or the control, which may not provide them with the experience they would like to achieve. This may be reduced by a crossover design, or by offering control participants the opportunity to undertake the intervention after completion of the follow up. Recruitment of a similar-sized sample of controls would provide a stronger design; however, far fewer controls were available. Finally, this initial study design would have been greatly improved by the use of a follow up survey, and the problem with so many interventions being that people's participation rapidly declines.
Therefore, follow-up will be undertaken in subsequent intervention studies to be performed. From participants' comments, it appears that the majority of those surveyed wished to continue outdoor swimming once the course had been completed either through formally joining clubs or meeting up with friends they made on the course. Consequently, the course provided them with skills, knowledge and contacts to potentially continue the activity upon the completion of the course.
In conclusion, novice outdoor swimmers had acute and chronic reductions in negative mood, increases in well-being and acute increases in positive mood. Controls' mood scores fluctuated and were similar at the start and end of the course, whereas well-being scores improved by the final session. Furthermore, tension scores peaked immediately in both swimmers and controls before the first outdoor swimming session. Nonetheless the swimmers improvement in mood and well-being scores were significantly greater than the controls.
Therefore, the hypothesis stating that there would be acute as well as longer term changes in mood and well-being as a consequence of participation in an outdoor swimming programme is accepted. The nature of this study does not provide mechanistic understanding; there are likely to be a number of explanations for the changes in mood and wellbeing in swimmers and controls that can be investigated in the future.