Ultra-endurance athletes’ use of fitness technologies: Filling a social void or fuelling an exercise addiction.

Dr Josephine Perry

Abstract

With a strong focus on finding solutions to the world’s obesity crisis one route which has been taken

is the development of the quantified self-movement utilising fitness technologies and social media

tools to increase adherence to physical activity. Early research is finding that this is proving to be

successful (Zhang, Brackbill, Yang, & Centola, 2016; Barwais & Cuddihy, 2015; Maher, Ferguson,

Vandelanotte & Olds, 2015; Goodyear, Kerner & Quennerstedt, 2017). A side effect is that the same

technologies are being used by those who are already highly motivated to exercise and this study

finds they are causing some to adhere to exercise so strongly that they are now at increased risk of

exercise addiction. The population studied in this research; 255 ultra-endurance athletes, were

found to have a risk of exercise addiction prevalence rate of 44.7% with those using a large number

of technologies in their sport having the highest risks. Six of those high technology using athletes at

risk of exercise addiction were interviewed. Five themes were identified as the athletes discussed

how they were not addicted to exercise, but to their sport, that they were all very individual in the

technologies they were using and why, that they were often seeking out an online community to

cope with the loneliness of their training and that the technologies they were using were helpful but

also causing harm and additional pressures.

Introduction

In 2017, 39% of the world’s population were estimated to be overweight (World Health

Organisation, 2017). Thousands of campaigns and strategies were focused on trying to reduce this

figure; usually based on education and tactics to limit energy intake and increase regular physical

activity (World Health Organisation, 2017). Many of these tactics involve technology; increasingly

through the ‘quantified-self’ movement, to help people track their physical activity, to stay

motivated and to engage with others also trying to improve their health. It is hoped these

technological tools will help individuals become educated and motivated towards adopting healthier

habits by recording and reporting information about their exercise, eating or sleeping behaviours

(Patel, Asch & Volpp, 2015).

1

Corporates have seen there is a real opportunity here and developed the quantified-self movement

into big business. The wearable tracking market in 2016 was estimated to be worth $61.7 billion. The

global technology player Nokia paid $191 million (Engadget, 2016) for device maker Withings so they

could gain a foothold in this market and in 2015 Under Armour paid $475 million for the fitness App

‘My Fitness Pal’ (Wall Street Journal, 2015). These apps are popular. In Nokia’s first six months in the

quantified-self market their health app was downloaded over one million times in the UK (BBC,

2017). These high cost purchases show the market is growing, and early research suggests it is

proving beneficial to their users.

Early research on the impact of these connected health technologies and networks has found

significant benefits particularly in affecting behavioural change for those new to exercise and

struggling to maintain motivation and adherence. Some of this motivation and adherence has been

found when individuals are using social networks to compare their activity levels to others (Zhang,

Brackbill, Yang, & Centola (2016) or when gadgets such as pedometers have been involved (Maher,

Ferguson, Vandelanotte and Olds (2015). The initial results are exciting with Barwais and Cuddihy

(2015) tracking the impact of an online personal activity monitor on reducing sedentary behavior

finding a significant decrease of 33% spent on sedentary activities and a significant increase in

energy expenditure and physical activity patterns after a four-week intervention. A study by

Goodyear, Kerner & Quennerstedt (2017) used the wearable tracking device, the Fitbit, giving them

to 100 high school students in the UK. They found the young people with the Fitbits increased their

physical activity not just due to the monitoring elements but also because peers were comparing

themselves with each other.

Three systematic reviews in this area are helping us understand more specifically which elements of

this quantified-self movement have the biggest impact on creating health behaviour change. They

have identified that the current most effective social networking sites for changing health

behaviours are Facebook, specialist health sites and then Twitter (Laranjo, Arguel, Neves & Lau

(2014), that there is a small positive effect of using connected health technologies, wearables and

social networks (Maher, Lewis, Ferrar & Vandelanotte, 2014) but that working purely through social

media (i.e. consuming rather than engaging) does not create significant changes in behaviour

(Williams, Hamm, Shulham & Hartling, 2014).

Many athletes are early adopters of technology and have also been using connected health

technologies to track their exercise, even though they are already completing far more hours of

2

exercise than minimum guidelines suggest. Those who are especially competitive may enjoy the

continual feedback they are able to get through quantified-self tools, expanding their opportunities

for comparison outside of competitions or races. If these tools are being found to increase the

amount of, and adherence to, physical activity in those who are not already motivated to exercise,

what will they do to highly motivated athletes who are already completing many hours of exercise? Is

there is a risk that the use of these technologies by those who are already seriously exercising could

increase their use of exercise to such an extent that they risk exercise addiction. Laud (2016) has

begun to consider this paradox and, with this massive increase in use, this seems a research area rife

for investigation.

A behavioural addiction, such as exercise addiction, where an individual tries to artificially change

their physical or mental state (Krivoschekov & Lushnikov, 2011) can often start out as beneficial (or

certainly neutral) to the individual but over time progresses to a state that is pathologically excessive

(Coom, Hausenblas & Freimuth, 2013). This is particularly obvious in exercise addiction where the

physical-effort and energy expenditure required for exercise can initially positively improve mental

and physical health (Warburton, Nicol & Bredin, 2006). Over time however its use can develop into

an obsessive and unhealthy preoccupation (Hamer & Karageorghis, 2007) where the exerciser uses

exercise to modify their mood, requires increasingly higher doses and gets frustrated and angry at

the thought of missing a session (Lichenstein, Christiansen, Bilenberg & Stoving, 2012). These three

elements often develop in stages until it creates a disorder (Weinstein & Weinstein, 2013). At this

point the athlete will see physiological changes during withdrawal (Antunes, Leite, Lee & de Mello,

2016) and can relapse when stopping. In short, they progressively lose self-control (Dakwar, 2015) of

their exercising.

The tipping point from when previously beneficial behaviour becomes an excessive and harmful

appears to be when the compulsion to exercise is prioritised over other parts of the athlete’s

lifestyle; harming their social relationships, work focus or family time and causing conflicts

(Lichtenstein, Larsen, Christiansen & Bredahl, 2014). These addicted exercisers are more likely to

exercise for intrinsic rewards, see exercise as a central component of their lives, feel deprived when

they are unable to exercise (Sachs, 1981) and those suffering from this often continue to exercise,

despite damaging effects such as injury, personal inconvenience, marital strain, interference with

work or reduced time for other activities (Landolfi, 2012) as well as lower self-satisfaction, poor

social behaviour and reduced vigor (Li, Nie & Ren, 2015).

3

The incomplete conceptual models of exercise addiction (Egorov & Szavo, 2013), the fact exercise

addiction is not listed in DSM-5 (APA, 2013) and the way that researchers often prefer to refer to

‘problematic exercise’ (Dakwar, 2015), excessive exercise (Berczik, Szabo, Griffiths, Kurimay, Kun,

Urban & Demetrovics, 2012), exercise dependence (Cockerill & Riddington, 1996) or compulsive

exercise (Dalle Grave, Calugi & Marchesini, 2008) instead of ‘addiction’ means there is a huge

amount of inconsistency within reported prevalence rates. Prevalence rates in the general public

(usually students) have been found to range from 2.5% in a UK study (Terry, Szabo & Griffiths, 2004)

to as high as 25.6% in a Canadian study (MacLaren & Best, 2010). Within athlete populations rates

have been found to be significantly higher. A study on Spanish body builders suggested the

prevalence rate was 26.7% (Salazar, Castro, Zambrano & Buitrago, 2012). A study from Blaydon and

Lindner (2002) found 30.4% of Hong Kong Triathletes had a risk of primary exercise addiction,

McNamara and McCabe (2012) found 34% of Australian elite athletes had this risk and 27% of ultra-

runners were also found to be at risk (Scheer & Hahn, 2014).

Weinstein and Weinstein (2016) have proposed that individuals more prone to developing an

exercise addiction may gravitate towards certain sports, endurance sports in particular. This is

supported by research in triathlon where Youngman & Simpson (2014) studied 1,285 triathletes and

suggested that 20% of these triathletes were at risk for exercise addiction. Additionally, an Australian

study (Magee, Buchanan, & Barrie, 2016) found that approximately 30% of Ironman participants

belonged to the ‘at-risk’ and ‘symptomatic’ profiles which could reflect maladaptive patterns of

exercise.

Elements within the type of exercise itself may also play some part in an athlete’s addiction risk. In

triathlon it has previously been found the longer the race distance, the higher the risk of addiction

and, as the number of weekly training hours increased, so did a triathlete’s risk of addiction

(Youngman & Simpson, 2014). When distance runners were researched it was found the more

competitive runners had the highest risk (Smith, Wright & Winrow, 2010). A number of other

individual factors found to increase risk of exercise addiction include age and a higher BMI

(McNamara & McCabe, 2012). The BMI element is important as disordered eating has been linked

closely with exercise addiction (Cunningham, Pearman & Brewerton, 2016) with 28% of the eating

disordered patients of Brewerton, Stellefson, Hibbs, Hodges and Cochrane (1995) exercising

compulsively and 48% of the anorexia patients researched by Klein, Bennett, Schebendach and

Walsh (2004) having a dependence on exercise. It has also been found that the environment in which

4

they trained (Allegre, Therme & Griffiths, 2007) and the social context of the training (Szabo, De la

Vega, Ruiz & Rivera, 2013) impact risk of addiction.

With a key element of addiction being tolerance and salience, it makes sense that sports which

require a huge amount of training, such as the ultra-endurance sports of marathons, ironman

triathlons and cycle sportives would be popular with those at risk of addiction. The variables

identified as increasing success in endurance sports such as a continual increase in training speed,

time and focus (Knechtle, Knechtle, Rosemann & Lepers, 2010) are similar variables identified in

those at risk of an exercise addiction. So, as an endurance athlete takes their sport more seriously

and continually increases the amount of training as the body adapts to it, perhaps this could start a

cycle of addiction, a risk identified by Scheer and Hahn (2014).

With the risk factors for exercise addiction to date being identified as; competing in ultra-endurance

sports, age, BMI, having an eating disorder, and having a personality high in traits of narcissism

(Bruno, Quattrone, Scimeca & Muscatello, 2014), extroversion and conscientiousness, (Andreassen,

Griffiths, Gjertsen, & Pallesen, 2013), excitement-seeking, perfectionism and achievement striving

(Lichenstein, Christiansen, Elklit, Stoving, 2013) it would be interesting to see if use of connected

health technologies can be added to these. To delve into this, the research set out two key questions:

1. Does high use of fitness technology increase the risk of exercise addiction?

2. How are those athletes at high risk of exercise addiction impacted by their use of fitness

technologies?

As ultra-distance endurance athletes not only have a potentially increased risks of exercise addiction

but also a reputation for early adoption of sports technologies and extensive ownership of sporting

‘gadgets’ they seem to be an ideal group to help us understand how the use of technology impacts

on an athlete’s risk of exercise addiction. Only one published study to date has looked at any

relationship between exercise addiction and technology. This study (Lejoyeux, Avril, Richoux,

Embouazza & Nivoli, 2008) assessed 300 adolescents in a French gym and found that those who had

exercise addiction spent more time on their computer (3.9 hours vs 2.4 hours per day) and more

time online than those without the addiction. This study ran years before many of today’s connected

health technologies were developed so it feels valuable to begin the process in much more depth to

understand all the ways in which athletes are moving towards using technology; so not just

wearables and apps but online trackers and reporting systems, online training plans, social media,

blog writing and forum engagement.

5

Methodology

To establish whether high use of fitness technology increases the risk of exercise addiction, a simple

quantitative approach, establishing sport focus, demographics, connected health technology use and

addiction risk was used. As behavioural addictions are ‘social constructions’ (Reinarmann, 2005)

their effects can often have a social basis and the patterns of use and impact can be shaped by the

social situations and the conditions surrounding the addict. These individual social environments and

the context based focus of an addiction means that qualitative research is a well-suited route to

really capture the lived experience of those with an addiction (Rhodes & Coomber, 2010) and to help

us explore any associations (Maher, 1997) so this approach was used for the second research

question; to understand how athletes at high risk of exercise addiction are impacted by their use of

fitness technologies.

Study 1

Participants

255 amateur athletes completed an online survey. Participants ranged in age from 19 to 70 years

(M=41.09). All participants took part in ultra-endurance sport (classified as: regularly running over

marathon distance, swimming over 5k, cycling over 100k or triathlon over half Ironman distance)

and had competed in at least one of these sports in the last 12 months. Participants were recruited

through Twitter, online endurance sport groups, Facebook and emailed to running, cycling and

triathlon clubs.

Measures

There are a number of measures used to understand a participant’s exercise addiction risk. The four

most used (and psychometrically tested) measures are the Obligatory Exercise Questionnaire (QED;

Ackard, Brehm & Steffenm 2002), Exercise Dependence Scale (EDS; Hausenblas & Symons Downs,

2002), the Exercise Dependence Questionnaire (EDQ; Ogden, Veale & Summers, 1997) and the

Exercise Addiction Inventory) (EAI; Terry, Szabo & Griffiths, 2004). Of all the scales, the EAI appealed

for this study as it is a short six-item tool which highlights the six common areas of exercise

addiction; salience, mood modification, tolerance, withdrawal, social conflict and relapse. Of

6

importance is that this tool captures the social conflict element which research has identified as

being a key way of highlighting when the positives of exercise are no longer being met and a

negative addiction is in place. As the measure consists of only six questions it is quick to use and has

high internal validity (Berczik, Griffiths, Szabo, Kurimay, Urban & Demetrovics, 2014). The Cronbach

alpha for the EAI is relatively high at 0.84. The measure is based on a five-point likert scale with the

scores added together to give a score out of 30. Those scoring over 24 are considered to be ‘at risk’

of an exercise addiction. A diagnosis of exercise addiction cannot be made in this way, nor should it

for research purposes without individualised support for impacted athletes, so this ‘at risk’ outcome

is the closest measurement we have for a study of this kind.

To measure technology use participants were asked to rate their use of nine types of connected

health technologies or social media that are often used by athletes: Tracking devices (GPS watch,

Online tracker, online diary), Social networking (Twitter, membership of an online forum), Dark

social (Facebook, WhatsApp Forum) or broadcasting (Blogging, Listening to Podcasts). This was rated

in a six-point likert scale from 0 (never use), 1 (use occasionally), 2 (use monthly), 3 (Use weekly), 4

(use every few days) and 5 (use every day). The total scores for all nine mediums were added to give

a total ‘technology usage’ score out of 45.

To understand the sport and training focus of the participants questions asked type of sports

competed in, their preferred sport, the amount of time spent training per week and how long they

have been competing in their sport. Participants were also asked to provide their age.

Procedure

Following ethical approval through the UCLAN Ethics committee potential participants were

contacted through online forums and social media based around endurance sports. Those agreeing

to participate completed the measures and questions through a 10-minute online questionnaire

from their home or work computer at a time which suited them. All participants answered the

questions in the same order following on-screen instructions.

Study 1 results

44.7% of ultra-endurance athletes within the study were shown to be at risk of exercise addiction.

Based on their current chosen sports this broke down into: Cycling: 39.58%, Triathlon: 46.07%,

Running: 44.26%. There were too few swimmers (5) in the study to separate out this group.

7

Using a linear regression we found that high use of fitness technology does increase the risk of

exercise addiction. There was a positive correlation between the level of use of fitness technology

(M=17.00, SD=6.25) and risk of exercise addition (M=22.66, SD=3.70), r=.21, p=.001. 4.6% of the

variance in risk of exercise addiction can be explained by the level of use of fitness technology (F 1,

254) = 12.17, p=.001.

More on the training preferences of the athletes and the technologies they use can be found in

Apprentice A1.

Study 2

Participants

Participants in study 1 were asked to indicate at the end of the study if they were interested in

participating in an in-depth interview study. 69 participants offered their involvement. 12 of these

athletes scored as being at risk of exercise addiction and also had high levels of technology use. Eight

of these participants were contacted and six agreed to participate in the interview.

Interview guide

With this being a new area of research and one which has yet to be widely theorised the focus

within the interviews was on discovery and learning, progressively focusing rather than matching a

hypothesis. In keeping with the inductive philosophy of research, an interview guide was created as

a base but acknowledged to be flexible and accepted that it would be developed as the interview

processes progressed. Based on this interview guide [Appendix A3] semi-structured interviews were

conducted with questions focused on sporting history, training habits, technology use, and the six

elements that impact on an athlete’s risk of exercise addiction.

Ethics

The participants in the qualitative study had all tested as highly at risk of an addiction, yet it is not an

addiction that would be seen by many as shameful or would marginalise them from their

communities. Unlike some other behavioural addictions, exercise is not illegal, its use at lower levels

is seen as incredibly beneficial, partaking in it is usually valued and the behaviour is rarely subject to

moral judgment. This meant that some of the usual ethical quandaries within behavioural addiction

8

research may have not been a concern for the participants but are important to maintain within

professional research. The key concepts of voluntary participation, minimising risk of harm,

providing informed consent and complete confidentially and having the proposal reviewed by an

ethics board (Miller, Carter & Hall, 2010) were all followed. Ethical issues specific to addiction; fully

informed consent, consideration of subject payment, rationality in choices and autonomy and

confidentiality and anonymity were also extensively considered and where relevant incorporated

into the methodological framework used.

Procedure and analysis

All participants were given information sheets and signed consent forms. Participants were all

interviewed over the telephone. The interviews lasted between 35 and 52 minutes (average 44

minutes). All interviews were sent to the participants after transcription and no changes were

requested.

The interviews were transcribed and analysed using Interpretive Phenomenology Analysis (IPA)

(Smith, 1996), a method allowing the researcher to develop themes and master themes to

understand in-depth the motivation and attitudes towards concepts and issues which can be difficult

to explain using quantitative data alone. IPA allows a homogenous group to be investigated, in this

case ultra-endurance athletes all at risk of exercise addiction and having high usage of technology for

their sport, to understand their lived experiences (Smith, 1996). The methodology allows the

researcher to gain an in-depth understanding of athletes with risk of exercise addiction and high

technology usage but does mean that there is no comparison to those with no risk of addiction or

low technology use. As a result, it is not possible to say that some attributes, attitudes or

characteristics are unique to those with a risk of exercise addiction but it does allow a rounded,

detailed description of this specific group of athletes, highlighting how they feel they behave in their

sporting environment, particularly when it comes to the way they are utilising technology and the

impact its use is having on them. Following IPA procedures, the transcripts were analysed for

recurring themes (Smith, 2011). The transcripts were read and possible broad themes were

identified. If these emergent themes were repeatedly found across and within interviews, they were

noted as recurrent themes. Groups of related recurrent themes were organised under a master

theme. The themes and master themes can be found in Appendix A4.

9

Study 2 results

Theme 1: Not addicted to exercise, but to their sport

All the athletes interviewed had score 24 or higher on the risk of addiction scale (which indicates

there may be a risk of addiction) and during their interviews they all were all asked questions

focused on the key elements which have been found to define exercise addiction. They all felt their

exercise was incredibly important to them, had all increased the amount of training they were doing

over time and all used exercise to improve their mood.

Kelly - The training is a stress reliever. I have quite a stressful job and it’s not a regular kind of 9:00 to 5:00, I find the discipline and the repetitiveness of the training helps me to switch off. I’m not a person who switches off very easily, but I find when I’m training it just focuses me over it.

The athletes struggled when they could not exercise, particularly with guilt and always got back into

it after time out with the same or more intensity.

Lesley – [when injured] I was generally quite miserable. I felt really down. There’d be times where I would literally put the key through the door, I’d be in tears, I’d go, “Oh my life is rubbish, I can’t even run. I’ve wasted a whole year. I spent eight hours at work, not doing anything because I don’t really enjoy it and now I can’t even go to a run.” I just moped a bit. Yes, I moped, mopey. Yes, feel generally like complete lack of achievement, like I’m going nowhere. The world is going far from me.

They all used sport to give their lives a focus and many said they felt aimless without it.

Grant - For me, it's a lot of mental, psychological. It really helps me having that kind of thing to focus on because I tend to be very aimless in life otherwise. I would say that's probably the biggest, mental one, for me. It actually makes me feel like I'm doing something with my life. It's not that I don't enjoy my work but I never saw work as something that I wanted to be the primary focus of my life. I would say exercise being my primary hobby has probably made it the most important thing in my life.

What seemed surprising and against the grain of what was expected though was that most of these

athletes were not exercising for the sake of it, they were exercising to be able to take part in their

sport. They all admitted that without a race to enter they would find it difficult to train. So it felt like

their addiction risk was not to exercise but to racing and their sport.

10

Jenny - I don't do the training because I enjoy training. I train to perform in the race, definitely. So I guess what keeps me competing is that feeling of racing and crossing the finish line really. I mean, obviously there are other goals in amongst that. Certainly, I don't train to train, I'm not one of those. But I do like it- don't get me wrong, I don't go out there every day and go, “Oh my God, I hate this,” but I can't see that I would do it to the extent that I do it if I didn't have the goal and a race to train for.

Behavioural dependences only become classified as addictions once they impact negatively on

someone’s life. For every interviewee it was clear that they had at least one difficult relationship

with either family or friends due to the amount of training they were doing.

Jenny - I did have a screaming row with my father over it on the phone... My father is probably one of the most competitive people I've ever met in my entire life and yet, he sees triathlon as an unhealthy obsession. My parents certainly see it as an unhealthy obsession and although they tell me, “Well done,” and, “We're proud of you,” it's certainly not something they want to talk to me about…At one point, I had an enormous row with my father and he just said, “Oh, you're just completely obsessed with this stupid sport,” …I put the phone down on him.

On a positive note however, it did seem that although it caused friction with loved ones, their

training was not replacing time spent with family and friends, it was replacing time they had

previously spent working. All these athletes had been working incredibly long hours before they had

got into sport and that had now taken a backseat, ironically giving them a better work-life balance.

Kelly - I reduced work. It’s literally been a direct drop. I see it as a good thing because I’m getting a lot more, physically, mentally fulfilled from the training than I would be from work. Although I like work, I thought that I didn’t have a work life balance. This is purely for me, it’s for nobody else.

Theme 2: There is no unified way in which athletes are choosing how to use technology

The athletes interviewed were using different types of technology for different purposes. The one

technology all were using were GPS watches (as did 92% of athletes in the quantitative research)

which they upload to sites like Garmin Connect or Strava to understand how they are performing

and whether they are improving.

Outside of the data tracking though it seems that the ways athletes are making their choices about

which ways to use and ‘consume’ their technologies may relate to their self-identity as an athlete.

11

Those who don’t feel like they are athletes are simply passively consuming information. They are

using their social media feeds as ways to gather information, take advice and learn more about their

sport. They are following athletes or sports organisations on twitter, reading athlete’s Facebook

chats, reading blogs, listening to Podcasts to see what other athletes are doing, to learn about races,

to read articles, to get advice, to see new training information and to feel part of their sport’s

community.

Grant - A lot of the time it's just interest about races. A lot of them post up little tidbits about races that are always quite interesting to read, whether it's just curiosity or because it's a race I'm thinking of maybe doing in the future. A lot of it is just kind of keeping up with news relevant to the sport, which, usually the quickest way to get that is through the people it directly affects.

Those self-identifying as an athlete where more proactive; often broadcasting their data, sharing it

by sending out tweets, starting Facebook discussions or writing a blog.

Katie - If anybody asks me for what I train, how I train, I’ll tell people. I happily share the advice I get with people who are not as experienced because that’s how I learned. I talk to other people in [race team] about their training. I sometimes enjoy comparing what other coaches do. We just compare sessions.

Beyond this some are also keen not just to consume, track, compare and broadcast their data but

additionally to participate with other athletes. They will be engaging with data feeds or social media

to discuss training, feedback on races or training sessions, comment on tracking data and engaging

with other members of their sports clubs.

Kelly - So, I would look at that and comment on that a bit, but more like if somebody said, “Oh this has happened to me,” and you know, you share your version of it, but I’m not actually saying “Here’s what I did today,” or anything like that.

There is also a range of ways that athletes are interacting with the data they produce through the

tracking technologies. Some, particularly those doing their sport to compete (i.e. race just to finish)

use it purely to compare to their previous data – not other people’s. Those who compete in races (to

beat previous times or beat others) use the data to benchmarking themselves against others. For

some this comparison can have a positive effect, with one athlete looking through her data to pull

out something positive from every session but for others they fid themselves self-judging each

session purely based on numbers, taking away wider exercise benefits such as enjoyment, feel or

flow.

12

Lesley - I feel like that for the bulk of my workouts, I spend it simply trying to chase the numbers on the screen. Ultimately, my sense of satisfaction on a run or ride is not from how I felt, but how I ranked on Strava! That sounds a bit sad doesn’t it? On one hand, it drives me to keep improving and training (a good thing I think), but on the other hand, it strips all of the other things that go along with exercise, and condenses it to the objective to run/ride the fastest, less of a good thing I think.

Theme 3: Community overriding loneliness

The athletes all said they would like to be able to train with others but, if they are to perform at their

best, they have to train on their own so they can train at the right level of intensity. This need to be

anti-social within their training can make their sport quite a lonely experience.

Grant - I'd say my general preference is I prefer to train with other people but when I'm getting into race-specific training or in certain sessions where I have a definite regimented training plan, I find it easier to do by myself so I don't have to worry about other people's mixed levels compared to mine and adjusting the session for that.

In addition to this, in almost all cases, as shown above, real life relationships had suffered due to the

athlete’s need to exercise. Due to these relationship issues and the need to train alone it seems

online communities (social networking, dark media or the engagement elements around their data

tracking) help to alleviate some of their loneliness. With training and racing taking up time previously

spent in work or having a social life, the athletes were finding newer, encouraging and more

accommodating relationships on line within social media and these were sometimes replacing real

world engagement.

Lesley - Just getting to know a group of people who will have similar thoughts and ambitions from what they want to achieve in sport. I would say that’s the main thing I’ve got out of it…I’d say we probably become, maybe it’s that feeling of being connected more via social media, simply because a lot of them – so I became a member of [city] Triathlon Group and a lot of their competitions and social events are all organised via Facebook and obviously everyone posts stuff up on Strava. I guess it’s been reinforced by that link on the social media, mainly Facebook. Essentially, if someone’s done a ride you might stick a comment down saying. “Good ride,” and give them the thumbs up and that’s been reciprocated. I guess it does help you feel more included in the group because everyone is keeping, not keeping tabs, that’s the wrong word, but everyone’s very much aware of what everyone is doing…I feel I don’t get left out I guess. Triathlon and running can be quite a solitary sport, and you don’t have the motivations of fellow team mates to encourage you to train… I guess Strava is filling that void to a certain extent.

13

All the athletes were clear that they enjoyed the support and friendships they got from this

community; either receiving praise for performing well in races or the entertainment they got when

doing very long indoor training sessions. Some of these friends that have been made have yet to be

met in real life.

Jenny - There are people that I only know through social media in the tri world, which is awesome. A bit surreal really but really, really awesome. And then there are people that I've met at races who I've known on social media for several years and then it's like, “Wow, it's amazing to actually meet you in person. Some of them [pro athletes] I've met post having interacted with them on social media and it's almost like you already know them… I have met so many different friends from all over the world.

With ultra-athletes in the quantitative study stating they train for over 10 hours a week then this

community can not only reduce some of the isolation felt but it is also enables them to feel more

part of their sport.

Kelly - I’d use [Facebook] a few times a day and it’s following trends, following posts, just being nosy looking for information and seeing what’s going on…I started looking at posts and things from professional athletes, just nosiness I guess and habit maybe. I guess it makes me feel more like part of the sport. I never really thought about it like that but yes, I guess it does.

A major risk with this reliance for support on the online sports community comes when an athlete

gets injured. It can increase the isolation they feel and prompt feelings of jealousy or despondency

about what they cannot do. If as a result they avoid social media they lose their community and the

support network they have built around them. They feel the loss of their community strongly and it

amplifies the negativity they are feeling towards their injury. If they are training as a coping

mechanism for other things (often stress or mental health issues) then not being able to train, and

losing all support mechanisms at the same time, could exacerbate the original issues. All athletes

were clear they felt left out and lonely when not doing well and can see how well all the others are

doing and another was beating himself up with guilt when over his friend’s Whatsapp group he can

see what his friends are doing in training but he is too injured to join in.

Lesley – Whilst I was injured, I kept getting notifications from Strava to say that so-and-so had broken my course record for various segments. There is nothing worse than not being able to run, and receiving a notification saying: ‘Uh-oh. So-and-so just broke your record’. It’s a very brutal reminder of inadequacy!

14

Theme 4: Technology is sticky to the point of harm

As would be expected, all these athletes at risk of addiction suggest they give 100% commitment to

their training and display strong levels of diligence to their sport. Not just to doing their activity but

doing it to the best of their abilities. This could be because it would be physically dangerous not to

train properly for ultra-distance races, but it also seems that these athletes have personality traits

where dedication features strongly. This dedication though means that they can get fixated by the

data.

Lesley - I don’t know how I’ve fallen into the trap but if it’s not on Strava it didn’t happen, I hate it. That’s the mentality… I still feel compelled to use Strava! It is almost an obsessive urge to review my splits and segment achievements, and rank myself against the athletic community.

The way each different technology has been designed is playing into that dedication and is

maximising each athlete’s fixation. Within online training diaries (such as Training Peaks) completed

sessions equal green boxes, missed sessions red boxes and this focuses an athlete’s mind to

complete their sessions.

Kelly - It’s more than validation from me, “Yes, it’s done now, check the box, it’s a task.” Although I enjoy it, I realise it sounds crazy. It’s that it’s complete, check the box, move onto the next on… It’s a task complete and it’s a good feeling. …I am very much a “I like to be in control” person…I am a bit of a control freak, I like to have a list of tasks, work my way through them, but I very much like to be in control of what I’m doing.

In online trackers (such as Strava) you can ‘win’ challenges or badges for fastest times and this

spreads now into online training games (such as Zwift) where you can be given upgrades to your

virtual bike or kit when you play / race more.

Katie - I used to be really Strava, possibly addicted… I used to go out and have QOMs. I might just look for the segment I wanted to have. Back in the U.S the route we did for our Monday night club ride, over the time, it was my goal to have the QOM for every single segment on that ride, and by the time I left the U.S I had QOMs every single segment.

These tools that capture, save and broadcast your data in a very visual way can be reassuring for the

athlete to help them to see where they have progressed, even when they are feeling doubtful or

have lost some self-confidence. It can give them a sense of accomplishment. This is played upon by

those running many of these technologies who have learnt to gamify their tools to offer online

15

achievements and rewards. This seems to be very effective in drawing in athletes already strong in

competitiveness, increasing adherence to usage, making the technologies very sticky.

Lesley - The thing about endurance athletes, and triathletes, is that they tend to have a slightly obsessive element to their character (myself included!). And with social media, there is always an article or a link to something telling you how to get faster, quicker, leaner, stronger etc… I think that this can tap into the insecurities of athletes and constantly reminds them of how much better they could be… I think that endurance athletes have a tendency to think that more is always better – more training, more miles, more speed etc. There is a danger of losing the focus of one’s own training and specific programme as you try to match what 20 other people (who are virtual strangers) are doing!

The risk here is that, as this athlete suggests, it is making them deviate from their own ‘real life’ goals

and not only reduces their chances of achieving them but also increases their risk of injury.

Jenny - That marathon where I broke down afterwards, I hit over 100k running in one of those weeks, which I've never done in my life before. I know that I came very close to breaking down with illness, and that was just before the taper. I was so determined I was going to hit 100k that I went and I ran when I probably shouldn't have run and I really got close to being very, very sick in the lead up to that race.

Theme 5: Technology is adding pressure

It was clear from all the athletes that in using technologies in their training they were adding in

additional pressures. Some of these pressures they found helpful to their performances. The brutal

honesty of the data, and the means to compare it to your own previous data and to other’s data,

means that athletes cannot gloss over how they are doing. It was raising, sometimes bluntly, their

self-awareness of their strengths and weaknesses. They have been forced to become more objective

about how they are performing and their fitness levels and are using it to help them self-reflect. The

technologies have also been adapted by athletes to help them train when they are returning from

burnout and have felt they couldn’t trust their own judgement.

Jenny - It's a heart rate variability app for training. I suppose what I let that guide is how hard I'm going to push in a training session or not….I certainly can override an enormous amount of fatigue and still push beyond where I thought I was capable of pushing, which is something I'm thrilled I have that ability because I know not everybody does have that. But I think that's a hazard as well…Nowadays, I can use the technology, say, “Oh, this is telling me that I'm really fatigued and I feel fatigued. Okay, I'm going to back off today and I'm going to go bit easier.”

16

While the objectivity given by a piece of technology here can be beneficial from a training

perspective it can also create some unintended issues. Continually using technology and reporting

back socially through technology can leave athletes feeling under additional pressures. Some

athletes were able to use this pressure and accountability in a helpful manner, using their natural

competitiveness to use the knowledge that others are watching as an extra impetus to train harder,

or to fight malaise if they are feeling unmotivated.

Grant - When I've been feeling like I shouldn't or wouldn't have been capable of doing a session, it's pushed me to do it… I use it as a way to push me, but I don't worry about this person training harder than me. I just worry about training as hard as I can.

However, that same pressure can push athletes to update sites in a time-consuming manner and

make add in a feeling of responsibility to their followers to keep up with it.

Grant - In terms of actually posting content I do feel a little bit more beholden to do that to upkeep some kind of image. Even though the parameters that I'm working in might change depending on life context, I do still feel some obligation to do so… I feel like it would be kind of dishonest for me to not post certain things on there…last year I was updating it very regularly, like once a week, and it cost me a lot of time...I try and post about once a month but it causes me quite a lot of stress trying to work out when I'm going to find the time to write it a lot of the time and also when I'm actually writing it, everyone is their own worst critic. But it is very rare that I actually post something and think, “Yes, I actually think that's a good piece of writing.” I guess that causes a bit of stress as well.

A further negative pressure is coming from the way athletes are think about how a session will look

to others, even before or while they are still doing the session.

Lesley – Saturday, it was a really good ride and it was good average, I think the average time is about 18mph, 6 hours, which I was fairly pleased with. I was thinking, “God, this is going to look really good when I upload this.” I know my brother keeps tabs on it... It’s like, “Oh yes, he’s going to be really impressed with this.” I do feel like, I guess that just also comes from knowing that I know I’ve exceeded my own, I’ve improved my own performance, but I know that is all going to be known by other people.

Not everyone is so open and honest though. Some were clear that they sanitised what they

broadcast and were less transparent, either by not putting missed sessions online or by promoting

false goals to protect their reputation if their real goal is missed.

Lesley - I went out for a longish ride after work on Wednesday (50 miles) and switched the Garmin on before I set off. The battery was dead so could not record my ride. Immediately, I was disappointed that it would not be recorded and uploaded onto Strava for all to see! But

17

then as I pedalled along, it was refreshing not to have my average speed flashing at me, urging me to go faster! I knew that neither myself or anyone else would see my split breakdown, and everything suddenly felt less pressured. It was a timely reminder that ultimately, I do these kinds of activities for fun, rather than to perform for others to see.

This protected themselves from a situation where the online comparisons were pushing the athletes

to question themselves or second guess their coaches or training plans. The athletes recognised

when they had done this it impacted on their training causing them to train when they should be

resting or push themselves harder than they should which in some cases had caused injury and

overtraining or just feeling disappointed with themselves.

Lesley - Knowing that it’s going on people’s view does probably encourage me to run faster….I’d run a little bit faster than probably what I should. Maybe that’s why I got injured, I don’t know. I have questioned that a lot actually, maybe I was running too hard all the time…This weekend, I went for a 10-mile run. It was only meant to be steady, as I am still coming back from injury. I averaged 4:30/km, it felt good, the sun shone, I ran along the Brecon canal, the birds were tweeting, and I finished with a cappuccino at a country pub. It was perfect. Then I got home, uploaded the run to Strava, and immediately saw that a couple of my friends and people that I follow had run a couple of kms further than me, and my brother had run at a higher pace than me. Immediately, it took the gloss off my run, and I felt like I should have run faster and further than the 10 miles. My efforts suddenly felt inadequate.

The quandary comes when the athlete becomes aware of all of these issues, and the pressure they

are feeling under but realises the way to protect themselves from this pressure is to not share

anything through technology; thus, reducing the positive elements of the pressure and losing the

sense of community that they so crave.

Discussion

The prevalence levels of ‘risk of exercise addiction’ in this study are higher than have been published

previously. This is not particularly surprising though as previous predictors of high risk have been

found to be time spent exercising (Kim & Shin, 2010), distances being trained for (Scheer & Hahn,

2014) and age (Allegre, Therme & Griffiths, 2007). Ultra-endurance athletes are required to train for

long periods so they are able to complete the distances they are training for and the nature of the

sports involved means that athletes ‘build up’ the distances they compete in over time so by the

time they compete in ultra-distance races they will be significantly older than athletes in other

sports. The average age of the athlete in this study was 41 with over nine years of training behind

18

them so have had a long period for their training and competing to have become ingrained into their

lives and to feel uncomfortable and irritable if they have to stop for any reason. Most other studies

of exercise addiction have focused on athletes who are far younger such as 24.2 years (Aidman &

Wollard, 2003) and 27.7 (Szabo & Solstad, 2016), giving them far less time to have trained in their

sport.

On top of this, and it is acknowledged that it is a clear weakness, the participants were contacted

over social media or technology (email) so the study would have attracted those with higher social

media use. As the study found that risk of exercise addiction in ultra-endurance athletes is higher in

those who have high technology use this could have had an impact on the overall prevalence levels

identified. However, study one, while helpful to get a general understanding of prevalence and to

identify any relationship between the risk of exercise addiction and use of technology, was mainly

designed as a recruitment tool for study two; to find athletes who did have a risk of addiction and

had high technology usage. This means that the bias in initial participant recruitment would not

create any imbalance in the qualitative research element.

A further difficulty comes from the fact that, to date, there are no widely used, validated tools

designed to measure technology use. In fact, the types of technologies considered within this study

are changing and developing so quickly that by the time any scale or measure had been designed

and tested it may well be out of date. There are also a number of terms used to describe different

technologies which means that both descriptions given by the researchers studying the issue and the

individual being studies may not always match and capturing the specific technologies and tools

used is a fluid and imprecise activity. What is clear is that the use of these technologies and the

social engagement elements alongside them will only grow. So, while researching the use of

connected technologies around the quantified self is inherently tricky, attempting to do so is a

valuable goal. The more effective designers are becoming at making them increase physical activity

adherence, the greater the impact they could have on those at risk of exercise addiction.

While there is no specific previous research to compare these results against it has previously been

found that the environment in which they trained (Allegre, Therme & Griffiths, 2007) and the social

context of the training (Szabo, De la Vega, Ruiz & Rivera (2013) impacts risk of exercise addiction and

these fitness technologies and tools certainly shape the virtual environment and social engagement

of the athletes using them. Although the strength of the link between technology and addiction

found in this study was fairly small it is important as fitness technologies, trackers and social media

19

are so pervasive in today’s society. While these tools clearly help some athletes train and perform

better, the athletes at risk of addiction who were interviewed described how these tools can create

problems by adding negative pressures, more stress, increasing injury risk, lowering potential

performance and simply removing some of the joy that these athletes take from their sport.

A key hypothesis of exercise addiction is that it is based on physiological need; with ideas focusing

on arousal (Thompson & Blanton, 1987), the reward systems activating dopamine or opioid release

systems (Boecker et al, 2008), endorphin release (Farrell, Gates, Maksud & Morgan, 1982),

thermogenic regulation (Morgan & O’Connor, 1988) or catecholamine regulation (Cousineau,

Ferguson, De Champlain, Gauthier, Cote & Bourassa, 1977). These biological theories do not support

the findings in this research though as it was not the physical exercise which the athletes were

finding themselves addicted to, but their specific sport and the competitive element within it. They

were clear they would be unlikely to continue with just the exercise without the threat or challenge

of races to train for. A further area of biological study has focused on the way it can reduce stress

and create a "runners high" which may lead to exercise addiction (Weinstein & Weinstein, 2013).

However, when this idea was tested by looking at cognitive performance, mental state, flow

experience, and brain cortical activity of 11 ultramarathon runners they found perceived physical

relaxation and flow state increased significantly after an hour of running but decreased after that

(Wollseiffen, Schneider, Martin, Kerhervé, Klein, & Solomon, 2016). This could explain why some

running would become addictive, but not ultra-running which lasts for over two hours for even the

fastest runners.

A further hypothesis of exercise addiction comes from the behaviourist perspective where some

actions are positively reinforced and others negatively reinforced. Here sport can be seen as a

positive reinforcer as increased training, effort and focus usually improves the individual’s

performance over time. The use of technology and social networking, allowing athletes to see the

results of their exercising all the time, will cement this behavioural reinforcement. McAuley, Wraith

and Duncan (1991), when researching triathletes found that increases in perceived efficacy (like

mastery, competence, and control) have been experienced post-exercise and these will become the

positive reinforcer (McAuley, Wraith & Duncan, 1991). Seeing these results in full digital colour, on

your phone, watch or laptop screen, whenever you want them will provide continue reinforcement

and could proof irresistible to athletes already at risk of exercise addiction. With Skinner (1989)

proposing that reinforcement was superior to punishment in altering behaviour it could be

20

suggested that any detrimental effect of over exercising would be less of a deterrent than the

motivation from the positive and negative reinforcement of the behaviour (Gupta & Gupta, 2002).

The most compelling theme from the qualitative research however focused on the importance of

community to athletes and that they are using these online communities to override the loneliness

inherent in their solitary sport. One psychosocial theory of addiction; the dislocation theory, has not

been applied to exercise to date but may have some merit in light of the views expressed by

participants within this study. Starting from the perspective that humans are social beings who need

to establish and maintain our place in society in order to develop well as a functioning adult it

suggests we have a vital need for belonging, alongside needs for individual autonomy and

achievement. These needs are met we are able to achieve psychosocial integration (Alexander,

2008) and when we lose it we suffer dislocation or a “poverty of the spirit” (Polanyi, 1944, p157).

Alexander states that addiction is a way of adapting to dislocation, where an addict tightly focuses

their lifestyle as an adaptive substitute for the wider social contact they are missing (Alexander,

2008). What is really interesting from the qualitative results is that technology seems to be being

used to rebuild a community that may be have been lost from an athlete’s real (physical) life.

Athletes within the study stated they interact with and learn from others on social media to the

point here they consider them friends; but sometimes they have yet to even met these people in

real life. They were clear that they enjoyed this psychosocial integration and it was making up for a

dislocation they were suffering from. It would be incredibly insightful to study this same subject but

within a strict framework around dislocation theory to understand if there is some circular notion of

dislocation prompting the exercise addiction and the use of technology and social media becoming a

hook from which the athlete is trying to re-establish some sense of psychosocial integration.

Whichever route is followed, a substantive, evidence based model of exercise addiction would be

beneficial in order to facilitate shortcuts in reducing the risk of exercise addiction. Perhaps one way

to start the facilitation of this would be some longitudinal research to understand how athletes are

adopting and using technology, and how their risk of addiction is developing over time as their

training requirements increase, and their use of connected health technologies, apps and social

networking changes. Whether longitudinal or not, more research is certainly needed, especially

within high risk groups. Exercise addiction sits as an outlier in addiction studies and, as discussed

above, a model of exercise addiction is still a way off. In the meantime, to reduce risk of exercise

addiction, it will be helpful to continue much of the research which focuses on identifying any

21

personality traits, personal risk elements or the exercise choices which increase risk of exercise

addiction.

With a low understanding of exercise addiction amongst many and often a ‘badge of honour’ feeling

for those who may be at risk, it is important to consider how athletes can be supported in reducing

their risk of addiction. The ‘badge of honour’ theme is widely spread over social media with regular

memes and tweets focusing on being tough enough, training through injury and sport being more

important than anything else. Yet the darker side of these comments is that those suffering from an

exercise addiction are having to cope with conflicts in their personal life and are more at risk of

eating disorder symptoms and report more bodily pain and injuries (Lichenstein, Christiansen, Elklit

& Stoving, 2013).

Many ultra-endurance athletes are self-coached. These athletes are vulnerable as they have no

outside perspectives or oversight to see when they are at risk of pushing themselves towards

addiction. These athletes may also be taking their advice from others online or finding online

competitions or activities to keep them motivated, which as this study has found could increase their

addiction risk. Perhaps one way to support these athletes is to use the specialist endurance sports

media to raise awareness of exercise addiction, the risk factors for it, the consequences of it and

how to approach making changes if necessary. This would also be beneficial for raising awareness

with professional practitioners such as coaches and sport psychology practitioners. An element of

this awareness raising could be to make coaches aware of the Exercise Addiction Inventory (Terry,

Szabo & Griffiths, 2004) so they are able to use it with the athletes they coach. It is a short and

simple measure which would require only limited instruction to use. Using this inventory at regular

intervals with an athlete as their exercise levels increase, if they are worried about excessive

commitment from the athlete or are noticing maladaptive patterns of exercise could be beneficial to

ensure they are not tipping towards addiction.

Finally, it will be important for treatment options to become more widely researched and available

for those with an exercise addiction. Currently CBT and motivational interviewing are suggested

(Lichtenstein, Hinze, Emborg, Thomsen & Hemmingsen, 2017) but the wider literature is thin on

ways to resolve exercise addiction and with 44.7% being at risk of addiction evidence based,

successful ways to reduce this figure will be incredibly valuable.

22

References

Ackard, D. M., Brehm, B. J., & Steffen, J. J. (2002). Exercise and eating disorders in college-aged women: Profiling excessive exercisers. Eating Disorders, 10(1), 31-47.

Alexander, B. K. (2008). The Globalisation of Addiction: A study in the poverty of the spirit. Oxford. Oxford University Press.

Allegre, B., Therme, P., & Griffiths, M. (2007). Individual Factors and the Context of Physical Activity in Exercise Dependence: A Prospective Study of ‘Ultra-Marathoners’. International Journal of Mental Health and Addiction, 5(3) 233–243.

Andreassen, C. S., Griffiths, M. D., Gjertsen, S. R., & Pallesen, S. (2013). The relationship between behavioral addictions and the five-factor model of personality. Journal of Behavioural Addictions, 2, 90-99.

Antunes, H. K. M., Leite, G. S. F., Lee, K. S., & de Mello, M. T. (2016). Exercise deprivation increases negative mood in exercise-addicted subjects and modifies their biochemical markers. Physiology & Behaviour, 156, 182-190.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (DSM-5®). American Psychiatric Pub.

Barwais, F. A., & Cuddihy, T. F. (2015). Empowering Sedentary Adults to Reduce Sedentary Behavior and Increase Physical Activity Levels and Energy Expenditure: A Pilot Study. International Journal of Environmental Research and Public Health 12, 414-427.

BBC (10th July 2017). Nokia regrets Withings backlash. http://www.bbc.co.uk/news/technology-40555752 (accessed 28/10/17)

Berczik, K., Griffiths, M., Szabo, A., Kurimay, T., Urban, R. & Demetrovics, Z. (2014). In Kenneth Paul Rosenberg & Laura Curtis Feder (Eds.). Behavioural Addictions: Criteria, Evidence and Treatment. London: Academic Press.

Berczik, K., Szabó, A., Griffiths, M. D., Kurimay, T., Kun, B., Urbán, R., & Demetrovics, Z. (2012). Exercise addiction: symptoms, diagnosis, epidemiology, and etiology. Substance use & misuse, 47(4), 403-417.

Blaydon, M. J., & Lindner, K. J. (2002). Eating disorders and exercise dependence in triathletes. Eating Disorders, 10(1), 49-60.

Boecker, H., Sprenger, T., Spilker, M. E., Henriksen, G., Koppenhoefer, M., Wagner, K. J., Vale, M., Berthele, A. & Tolle, T. R. (2008). The Runner's High: Opioidergic Mechanisms in the Human Brain. Cerebal Cortex, 18(11) 2523-2531.

Brewerton, T., Stellefson R., Hibbs N., Hodges E., & Cochrane C. (1995). Comparison of eating disorder patients with and without compulsive exercising., International Journal of Eating Disorders, 17, 413-416.

Bruno, A., Quattrone, D., Scimeca, G., Cicciarelli, C., Romeo, V. M., Pandolfo, G. Zoccali, R. A. & Muscatello, M. R. A. (2014). Unravelling Exercise Addiction: The Role of Narcissism and Self-Esteem. Journal of Addiction,

Cockerill, I. M. & Riddlington, M. E. (1996). Exercise dependence and associated disorders: A review. Counselling Psychology Quarterly, 9(2), 119-129.

23

Cook, B., Hausenblas, H., & Freimuth, M. (2014). Exercise addiction and compulsive exercising: Relationship to eating disorders, substance use disorders, and addictive disorders. In Eating disorders, addictions and substance use disorders (pp. 127-144). Springer Berlin Heidelberg.

Cousineau, D., Ferguson, R. J., De Champlain, J., Gauthier, P., Cote, P., & Bourassa, M. (1977). Catecholamines in coronary sinus during exercise in man before and after training. Journal of Applied Physiology, 43(5), 801-806.

Cunningham, H. E., Pearman, S., & Brewerton, T. D. (2016). Conceptualizing primary and secondary pathological exercise using available measures of excessive exercise. International Journal of Eating Disorders, 49(8), 778-792.

Dakwar, E. (2015). Problematic exercise: A case of Alien feet. In Michael. S. Ascher & Petros Levounis, The Behavioural Addictions. Arlington, VA: American Psychological Association.

Dalle Grave, R. Calugi, S. & Marchesini, G. (2008). Compulsive exercise to control shape or weight in eating disorders: Prevalence, associated features and treatment outcome. Comprehensive Psychiatry, 49(4), 346-352.

Egorov, A. Y. & Szavo, A. (2013). The exercise paradox: An interactional model for a clearer conceptualization of exercise addiction. Journal of Behavioural Addictions, 2(4) 199-208.

Engadget (2016). https://www.engadget.com/2016/04/26/nokia-buys-withings/ (as accessed 25/10/17)

Farrell, P. A., Gates, W. K., Maksud, M. G., & Morgan, W. P. (1982). Increases in plasma beta-endorphin/beta-lipotropin immunoreactivity after treadmill running in humans. Journal of Applied Physiology, 52(5), 1245-1249.

Goodyear, V. A., Kerner, C., & Quennerstedt, M. (2017). Young people’s use of wearable healthy lifestyle technologies; surveillance, self-surveillance & resistance. Sport, Education and Society, 1-14.

Greenberg, J. L., Lewis, S. E., & Dodd, D. K. (1999). Overlapping addictions and self-esteem among college men and women. Addictive behaviours, 24(4), 565-571.

Gulker, M. G., Laskis, T. A., & Kuba, S. A. (2001). Do excessive exercisers have a higher rate of obsessive-compulsive symptomatology?. Psychology, health & medicine, 6(4), 387-398.

Gupta, D. S. & Gupta, D. M. (2002). Psychology for Psychiatrists. London: Whurr Publishers.

Hamer, M. & Karageorghis, C. I. (2007). Psychobiological Mechanisms of Exercise Dependence. Sports Medicine, 37(6) 477-484.

Hausenblas, H. A., & Downs, D. S. (2002). How much is too much? The development and validation of the exercise dependence scale. Psychology and health, 17(4), 387-404.

Knechtle, B., Duff, B., Schulze, I., Rosemann, T. & Senn, O. (2009). Anthropometry and pre-race experience of finishers and non finishers in a multistage ultra-endurance run. Perceptual and motor skills, 109(1) 105-118.

Knechtle, B., Knechtle, P., Rosemann, T. & Lepers, R. (2010). Predictor variables for a 100-km race time in male ultra-marathoners. Perceptual motor skills, 111(3) 681-693.

Krivoschekov, S. G. & Lushnikov, O. N. (2011). Psychophysiology of sports addictions (exercise addiction) . Human Physiology, 37(4) 135-140.

Landolfi, E., (2013) Exercise Addiction. Sport Medicine, 43(2) 111-119.

Laranjo, L., Arguel, A., Neves, A. L., Gallagher, A. M., Kaplan, R., Mortimer, N., ... & Lau, A. Y. (2014). The influence of social networking sites on health behavior change: a systematic review and meta-analysis. Journal of the American Medical Informatics Association, 22(1), 243-256.

24

Laud, G. (2016). Conference Paper: The Dark Side of Connected Health Technology: How Connectivity Creates Exercise Addiction among Conscientious Users. Conference: North American ACR.

Lejoyeux, M., Avril, M., Richoux, C., Embouazza, H., & Nivoli, F. (2008). Prevalence of exercise dependence and other behavioral addictions among clients of a Parisian fitness room. Comprehensive Psychiatry, 49(4), 353-358.

Li, M., Nie, J., & Ren, Y. (2015). Effects of exercise dependence on psychological health of Chinese college students. Psychiatria Danubina, 27(4) 413-419.

Lichenstein, M. B., Christiansen, E., Bilenberg, N., Stoving, R. K. (2012). Validation of the exercise addiction inventory in a Danish sport context. Scandinavian Journal of Medicine and Science in Sports, 24(2) 447-453.

Lichtenstein, M. B., Christiansen, E., Elklit, A., Stoving, R. (2014). Exercise addiction: A study of eating disorder symptoms, quality of life, personality traits and attachment styles. Psychiatry Res, 215(2) 410-416.

Lichtenstein, M. B., Larsen, K. S., Christiansen, E., & Bredahl, T. G. V. (2014). Exercise addiction in team sport and individual sport: Prevalences and validation of the exercise addiction inventory. Addiction Research and Theory, 24(2) 447-453.

Lichtenstein, M., Hinze, C. J., Emborg, B., Thomsen, F., & Hemmingsen, S. D. (2017). Exercise addiction: Links, risks and challenges faced. European Psychiatry, 41, S868-S869.

MacLaren, V. V., & Best, L. A. (2010). Non-suicidal self-injury, potentially addictive behaviours, and the Five Factor Model in undergraduates. Personality and Individual Differences, 49(5), 521-525.

Magee, C. A., Buchanan, I., & Barrie, L. (2016). Profiles of exercise dependence symptoms in Ironman participants. Psychology of Sport and Exercise, 24. 48-55.

Maher, C., Ferguson, M., Vandelanotte, C., Plotnikoff, R., De Bourdeaudhuij, I., Thomas, S., ... & Olds, T. (2015). A web-based, social networking physical activity intervention for insufficiently active adults delivered via Facebook app: randomized controlled trial. Journal of medical Internet research, 17(7).

Maher, C., Lewis, L., Ferrar, K. & Vandelanotte, C. (2014). Are Health Behavior Change Interventions That Use Online Social Networks Effective? A Systematic Review. Journal of Medical Internet Research 16(2) 40

Maher, L. (1997). Sexed Work: Gender, Race and Resistance in a Brooklyn Drug Market. Oxford: Clarendon Press.

McAuley E., Wraith S., Duncan T., (1991). Self-efficacy, perceptions of success, and intrinsic motivation for exercise. Journal of Applied Social Psychology, 21(2) 139-155.

McNamara, J. & McCabe, M. P. (2012) Striving for success or addiction? Exercise dependence among elite Australian athletes. Journal of Sports Sciences, 30(8), 755-766.

Miller, P. G., Carter, A., & Hall, W. (2010). Ethical issues in alcohol, other drugs and addiction-related research. In Peter. G. Miller, John Strang and Peter M Miller (Eds.). Addiction Research Methods. Chichester, UK: John Wiley & Sons Ltd.

Morgan, W. P. & O’Connor, P. J. (1988). Exercise and mental health. P99-121. In RK Dishman (Ed.) Exercise adherence: Its impact on public health Champaign, IL: Human Kinetics.

Ogden, J., Veale, D., & Summers, Z. (1997). The development and validation of the Exercise Dependence Questionnaire. Addiction Research, 5(4), 343-355.

Patel, M. S., Asch, D., A. & Volpp, K. G. (2015). Wearable Devices as Facilitators, Not Drivers, of Health Behavior Change. The Journal of the American Medical Association 313 (5), 459-460

25

Polanyi, K. (1944). The great transformation: the political and economic origins of our times. Boston, MA: Beacon.

Reinarmann, C. (2005). Addiction as an accomplishment: the discursive construction of disease. Addiction Research and Theory, 13, 307-320.

Rhodes, T., & Coomber, R. (2010). Qualitative methods and theory in addictions research. In Peter. G. Miller, John Strang and Peter M Miller (Eds). Addiction Research Methods. Chichester, UK: John Wiley & Sons Ltd.

Sachs. M. L. (1981) Running addiction. In Sacks, M., Sachs, M. (eds) Psychology of Running. p116–126. Human Kinetics, Champaign, ILL.

Salazar, J. A. A., Castro, S. G., Zambrano, C. R., & Buitrago, S. V. (2012). Levels of body exercise addiction in bodybuilders. Revista Iberoamericana de Psicologia del Ejercicio y el Deporte

Scheer, V. & Hahn, S. (2014). Conference Paper: Are ultramarathon runners exercise addicts? German Association of Sports Medicine; 45th Sports Medicine Congress Sept 2014.

Skinner, B. F. (1989) The Origins of Cognitive Thought. Recent Issues in the Analysis of Behaviour, Merrill Publishing Company.

Smith, D., Wright, C., & Winrow, D (2010). Exercise dependence and social physique anxiety in competitive and non competitive runners.‐ International Journal of Sport and Exercise Psychology, 8(1) 61-69.

Smith, J. A. (1996). Beyond the divide between cognition and discourse: Using interpretative phenomenological analysis in health psychology. Psychology and health, 11(2), 261-271.

Smith, J. A. (2011). Evaluating the contribution of interpretative phenomenological analysis. Health psychology review, 5(1), 9-27.

Szabo, A., De la Vega, R., Ruiz, R., & Rivera, O. (2013). Exercise addiction in Spanish athletes: Investigation of the roles of gender, social context and level of involvement . Journal of Behaviour Addiction, 2(4) 249-252.

Terry, A., Szabo, A., & Griffiths, M. (2004). The exercise addiction inventory: A new brief screening tool. Addiction research and theory, 12(5), 489-499.

Thompson, J. K., & Blanton, P. (1987). Energy conservation and exercise dependence: a sympathetic arousal hypothesis. Medicine & Science in Sports & Exercise.

Wall Street Journal (2015). https://www.wsj.com/articles/under-armour-to-acquire-myfitnesspal-for-475-million-1423086478 (accessed on 25/10/17)

Warburton, D. E., Nicol, C. W. & Bredin, S. S. (2006). Health benefits of physical activity: The evidence. CMAJ, 174(6), 801-809.

Weinstein, A. & Weinstein, Y. (2013). Exercise Addiction- Diagnosis, Bio-Psychological Mechanisms and Treatment Issues . Current Pharmaceutical design, 20(25) 4062-4069.

Weinstein, A. & Weinstein, Y. (2016). Exercise Addiction: Diasgnosis, Psychobiological Mechanisms and Treatment. In Nancy M. Petry (Ed). Behavioural addictions: DSM-5 and beyond. Oxford, UK: Oxford University Press.

Williams, G., Hamm, M. P., Shulham, J., & Hartling, L. (2014). Social media interventions for diet and exercise behaviours: A systematic review and meta-analysis of randomised controlled trials. British Medical Journal. Open 4 (2) 003926

Wollseiffen, P., Schneider, S., Martin, L. A., Kerhervé, H. A., Klein, T. & Solomon, C. (2016). The effect of 6 hours of running on brain activity, mood, and cognitive performance. 234(7) 1829-1836.

26

World Health Organisation (2017) http://www.who.int/mediacentre/factsheets/fs311/en/ (accessed on 25/10/17)

Youngman, J., & Simpson, D. (2014). Risk for Exercise Addiction: A Comparison of Triathletes Training for Sprint-, Olympic-, Half-Ironman-, and Ironman-Distance Triathlons. Journal of Clinical Sport Psychology, 8, 19-37.

Zhang, J., Brackbill, D., Yang, S., Becker, J., Herbert, N., & Centola, D. (2016). Support or competition? How online social networks increase physical activity: A randomized controlled trial. Preventive medicine reports, 4, 453-458.

Appendix

A1: The training and technology preferences of athletes in the quantitative study

Current preferred sport Running (44.3%)Triathlon (34.9%)Cycling (18.8%)Swimming (2%)

Number of years competing in current sport 9.25 yearsNumber of hours per week spent training 10.28 hoursAverage tech score - 9 types of technology scored between 0 (never use) and 5 (use daily)

19.96

Average exercise addiction score - scores between 6-30 22.66Average number of athletes using technologies at least once per week

GPS watch or tracker (92.16%)Online tracker (84.31%)Facebook (70.20%)Twitter (39.61%)Posting on forum (35.29%)Online training diary (34.51%)Being in WhatsApp group (27.84%) Listening to sports podcasts (18.43%)Blogging or updating website (6.67%)

27

A2: The training and technology preferences of athletes in the qualitative study

Name (Pseudonym)

Age Sport Length of time in sport

Hours training a week

Addiction score (out of 30)

Technologies used

Lesley 30 Triathlon 1 10 24 Strava, Facebook, a GPS watch, online tracker, club forum, Facebook messenger

Charles 70 Running 11 6 24 Twitter, Facebook, GPS watch, online tracker, forums, Whatsapp groups

Grant 27 Triathlon 2 12 27 Twitter, Facebook, a GPS watch, an online tracker, an online forum and an online diary

Jenny 43 Triathlon 13 15 27 Twitter, Facebook, GPS watch, online training diary, online tracking, podcasts

Katie 46 Cycling 6 20 25 Twitter, Facebook, GPS watch, online tracker, forums, Whatsapp groups

Kelly 42 Triathlete 2 12 25 Twitter, Facebook, GPS watch, online tracker, online training diary, forum

A3: Interview guide

Technology use: XX/45Addiction Risk: XX/30

Introduction From your survey answers it seems you focus on [sport]. Is that right? What benefits do you get from it? What keeps you competing?

Training practices You mention in the survey that you train about XX hours a week. Is that right? Do you enjoy the training? Would you still train if you were not competing? Do you prefer to train alone or with other people? If you were not training for [sport] how would you spend that time?

Technology used You mentioned you use Twitter, Facebook, a GPS watch, an online tracker, an online forum and

an online diary. What is your favourite piece of technology that you use for your sport? What makes you like it so much? Do you find any downsides with it?

28

Do you ever find the need to ‘report back’ through your technology means you are more likely to exercise?

Do you ever feel bad if you miss a session and have to report that through your technology? Do you follow other athletes on facebook? If so, how do you feel when you see them post about their training? Does seeing other people posting about their training increase your adherence to your own

exercise?

Salience How important is exercise to you? Can you describe that feeling? What other priorities do you have in your life? Would you be able to say where exercise fits in among these other priorities?

Conflicts Have you ever fallen out with anyone due to your exercising? Would you be able to tell me more about that?

Mood enhancement Do you ever use exercise as a way of changing your mood How does exercise change your mood? How would you describe your mood when you can’t exercise? Do you have any technological tools you rely on to get you exercising?

Dosage Have you found yourself doing more exercise over time? Why would you say that is? Has increasing the amount of exercise you do caused you any problems elsewhere in your life? Do you look back through your blogs, social media posts or online trackers to see what exercise

you have done previously? Do you compare what you are doing now to what you have done previously.

Withdrawal How do you feel when you can’t exercise? Do you ever feel guilty if you can’t exercise? Could you explain why?

Relapse Have you ever stopped your sport for a significant amount of time? Why was that? If for injury: how long did it take you to build up again? Did you build up at the speed you were

advised to? If not injury: Why was that? How did you feel about that? Could you see yourself stopping again?

Conclusion Is there anything you were expecting me to ask that I didn’t? Do you have any questions for me about the research?

29

A4: IPA research themes

30

Not addicted to exercise but their sport

Improved mood

Felt guilt when not training

Aimless without sport

Improved work-life balance

No unified way athletes choosing how to use technology

Personal tracking

Consumption

Passively

Participatory

Broadcasting

Comparison

Compete

Complete

Community overrides the loneliness

Train alone

Support

Loneliness

Friendships

Injury

Stickiness of the technology becoming harmful

Dedication

Fixation

Gamifaction

Deviation

Injury risk

Poor performance

Technology becomes a pressure

Positive pressures

Accountability

Honesty

Self-awareness & insight

Negative pressures

Time consuming

Comparisons

Sanitisation

Disappointment