“He’d Be Happier if He Wasn’t Chonky” – Qualitatively Exploring Canine Obesity Perceptions Using YouTube™ and Discussion Fora

ABSTRACT Canine obesity is one of the top welfare problems of pet dogs. Owners are often unable to successfully recognize and manage their dog’s condition, even with assistance from veterinarians. The aim of this exploratory study was to appraise people’s perceptions, attitudes, and behaviors relating to canine obesity and weight management by analyzing comments made in online public fora and about online video clips. Data comprised 450 posts on 15 threads related to canine obesity from online discussion fora (www.petforums.co.uk, www.reddit.com, and www.mumsnet.com) and 637 comments posted about five videos published online (www.youtube.com). These fora sites either had a dedicated topic area or were entirely for discussions regarding pets. Threads and videos chosen represented a diversity of obesity-related topics, dog breeds, and a range of overweight severities. Data were anonymized and analyzed using thematic analysis. Four key themes emerged: (1) Balancing conflicting responsibilities – Individuals appeared to balance their responsibility in providing their dog with happiness, health, and love, and differences in emphasis placed on these impacted feeding habits and weight management; (2) Need vs. greed – Individuals felt compelled to alleviate perceived hunger in their dog, which made sticking to reduced food diets difficult for some; (3) Minimizing – Individuals varied in the extent to which they perceived excess body fat to be problematic, and language used to describe their dog’s body changed when excess body fat was seen as an issue; (4) Control – Individuals’ perceived control over their dog’s body condition and food intake varied hugely, with some owners believing they had little-to-no control. Whilst such publicly available data need to be interpreted with caution, due to self-selection bias, this study provides valuable insight into factors that impact feeding practices and could impact compliance with weight-reduction programs. These findings can be incorporated into future research and behavior-change initiatives to increase engagement and compliance.

It is estimated that over half of the UK's pet dogs are overweight or have obesity (German et al., 2018), and a similar prevalence is also reported in studies from Europe, Australia, USA, Brazil, and China (Mao et al., 2013;Muñoz-Prieto et al., 2018; Association for Pet Obesity Prevention, 2019; Porsani et al., 2020;Rohlf et al., 2010). Not only does the prevalence of obesity appear to be increasing (Banfield., 2019;German et al., 2018), but more cases are developing severe obesity, whereby they exceed the maximum score on the widely used 9-point Body Condition Scoring system, which equates to ∼40% above ideal weight (German et al., 2021;Laflamme, 1997). Canine obesity is associated with a range of comorbidities, including: diabetes mellitus, hypothyroidism, hypertension, cardiac disease, respiratory diseases, neoplasia, osteoarthritis, and cruciate ligament disease (German, 2006). Canine obesity is also associated with decreased life expectancy (Salt et al., 2019) and quality of life (German et al., 2012). Successful weight loss is associated with improved quality of life (German et al., 2012), mobility (Marshall et al., 2009), and insulin sensitivity (German et al., 2009). Despite weight reduction having numerous benefits, amongst those that start a weight-loss program, almost half do not reach the target weight and, of those that do, many will relapse (Carciofi et al., 2005;German et al., 2007;German et al., 2012). Owners often underestimate the size of their own pet, even when shown how to use a Body Condition Scoring chart (Eastland-Jones et al., 2014). Despite the high prevalence of obesity, a report by the Pet Food Manufacturers Association found that only 8% of owners believed their pet needed to lose weight (Pet Food Manufacturers Association, 2019). This may contribute to the finding that when owners have been advised by a veterinarian that their dog is overweight, 50% disagree and maintain that their dog is the ideal weight (White et al., 2011).
Owners may find it hard to acknowledge obesity in their own pets owing to the portrayal of canine obesity in both veterinary articles and main-stream media which focusses on individual-level solutions, putting the onus on owners and suggesting that their dog's obesity is their own personal failing rather than a systematic issue (Degeling et al., 2011;Degeling & Rock, 2012). One study found veterinarians rated owner responsibility as the largest contributing factor to canine obesity and they felt the owners were to blame (Pearl et al., 2020). Furthermore, it has been found that despite veterinarians being the primary sources of health advice for dog owners (Kuhl et al., 2021), they rarely discuss weight management with owners (German & Morgan, 2008;Rolph et al., 2014), a common concern being offending owners (Aldewereld et al., 2021;Cairns-Haylor & Fordyce, 2017). When veterinarians discuss weight with owners, they vary in the depth of nutritional history taken and the detail provided about prevention and treatment (Phillips et al., 2017).
From the existing quantitative research, risk factors for canine obesity include dogspecific factors such as increased age, being neutered, being female, and being from certain breeds (e.g., Labrador retriever, Pug, and Beagle) (Courcier et al., 2010;Edney & Smith, 1986;Pegram et al., 2021). Owner factors are also associated with canine obesity, including increased age, lower socio-economic status, greater BMI, and permissive ownership styles, where owners are highly responsive to their dog's needs and expect relatively little from them (Courcier et al., 2010;Edney & Smith, 1986;Herwijnen et al., 2020). As valuable as these quantitative research insights are, they do not explain why obesity management and prevention is so challenging for owners and what the solutions might be.
Qualitative research methods are designed to explore the intricacies of everyday behavior and the reasons why people act in the way they do (Fitzpatrick & Boulton, 1996). These methods can be useful for understanding social phenomena in natural settings; notably, how people interpret health messages in relation to their own lives (Pope & Mays, 1995). Insights from qualitative forum and focus-group analysis include that parents understand the importance of prevention and interventions on childhood obesity but find it difficult to implement these messages in everyday life (Watkins & Jones, 2015). Such research has highlighted the importance of health professionals focusing on providing innovative, tailored support to parents to help them implement prevention and intervention strategies, rather than repeating messaging about the concerns of obesity (Appleton et al., 2017;Watkins & Jones, 2015). There have been limited qualitative studies investigating canine obesity and weight management. Studies have been limited to surveys (Cairns-Haylor & Fordyce, 2017;Wainwright et al., 2022;White et al., 2011) and focus groups (Downes et al., 2017) or have focused on veterinarians' experiences of diagnosing obesity (Aldewereld et al., 2021) or advising owners to increase dog walking (Burns et al., 2018).
In previous qualitative research studies, investigators successfully analyzed publicly available online data about various health topics, including fertility (Hanna & Gough, 2016), strokes (Jamison et al., 2018), and obesity in humans (Willmer & Salzmann-Erikson, 2018). These studies have led to an increased understanding of individual attitudes toward and perceptions of diagnosis and treatment options, highlighting areas where more information or support is needed from stakeholders. A recent exploration of equine obesity recognition and management used various qualitative methods, including analyzing discussion fora, indepth-interviews, and focus groups, providing insights into why owners struggled to recognize their horse's excess weight and found it difficult to implement changes to manage their horses' weight (Furtado et al., 2021). Such insights, which are only possible by using qualitative methods, led to the development of a decision-making guide to empower owners in creating tailored weight-management strategies. In another study, that analyzed data relating to obesity from various online platforms (including Twitter, Facebook, and fora), personal opinions were more commonly expressed in fora, as struggled with weight reduction and seeking direct advice from others (Chou et al., 2014). Comments posted in response to videos published online (www.youtube.com) also provided an opportunity to analyze the attitudes and perceptions of individuals regarding dog bites and attribution of blame (Owczarczak-Garstecka et al., 2018). Therefore, the aim of the current exploratory study was to explore perceptions, attitudes, and behaviors regarding canine obesity and weight management using comments on online fora and about videos of dogs posted online.

Methods
Discussion threads were collected from three online discussion fora websites, each with at least 200,000 registered users and millions of non-registered visitors each year (www. petforums.co.uk, www.reddit.com, and www.mumsnet.com). These fora sites were either entirely dedicated to discussions regarding pets or had a specific topic area dedicated to pets and were frequently used by pet-dog owners, typically with several new posts per day related to pet dogs. Fora were found by searching "dog discussion forum" using an internet search engine. Criteria for their use were that they must be searchable and open access. Relevant discussion threads within these fora and videos published online (www.youtube.com) were identified using the following search terms: "Dog overweight," "Dog obese," "Dog weight," "Dog fat," "Dog chubby," "Dog diet," "Dog hungry," and "Dog begging," There were 127 fora threads that matched these searches and 131 videos. Fifteen discussion threads and comments on five videos were purposively selected for analysis. As qualitative research focuses on understanding individuals lived experiences, the sampling design is not random, but seeks to include the most appropriate data sources for answering the research question (Etikan et al., 2016;Johnson et al., 2020). A best practice sampling method for qualitative research is purposive sampling; this involves researchers selecting information-rich cases for analysis to make efficient use of resources (Etikan et al., 2016;Johnson et al., 2020). In the current study, purposive sampling was used to select forum threads and videos that reflected a diversity of topics including dog breeds, overweight severities, and rescued or owned dogs. Other criteria for their selection were that the posts must be from within the previous 6 years (2015-2020) and the discussion must be relevant to canine obesity (see Table 1 for details about the selected threads and videos). All comments and posts in each forum thread were included in the analysis, totaling 450 fora comments. This was also the case for two videos where there were <150 comments. Given that there were several hundred comments about the remaining three videos, the 150 most recent comments were analyzed, totaling 637 comments across the five videos.
Comments were analyzed to gain insight into participants' perceptions, attitudes, and behaviors relating to canine obesity, weight management, feeding practices, and healthy lifestyles for dogs. Analyzing forum data and online comments was chosen as this was explorative research; we did not want to narrow the data based on existing assumptions, so we sought a collection method that minimized the researcher's influence and the data were analyzed inductively to avoid constricting analysis by imposing existing ideas or frameworks onto the analysis (Braun et al., 2017;Clarke & Braun, 2014;Jowett, 2015). Thematic analysis was deemed to be the most suitable method of data analysis because, through the process of systematic data coding, common themes around individual's perceptions, attitudes, and behaviors regarding canine obesity could be developed and summarized. This form of analysis is also useful for identifying the salient issues and typical responses of individuals in a particular group, and it helps with developing interventions that are attuned to this context (Degeling & Rock, 2020;Green & Thorogood, 2018). To increase rigor and further minimize the risk of the researcher's positionality impacting the analysis, the themes developed were discussed with the wider research team during regular meetings. This allowed for triangulation of the data, to increase understanding and aid with "tussling with the data" to develop an analysis that best answered the research question (Braun & Clarke, 2016).
Data analysis was conducted using NVivo10 software (QSR International) after removing any identifying details (usernames, locations, pet names). First, descriptive codes were applied to individual sentences, phrases, or words that summarized the meaning. As more data were added to these codes, they were refined and grouped together in analytical themes (Clarke & Braun, 2014). As the themes were refined, the links between themes An owner that noticed their dog was overweight asked for advice on weight reduction. The responses included advice on weight reduction and management and a discussion about how other owners had struggled to keep their own dog a healthy weight.

Mumsnet F2
A concerned owner posted a photo of their puppy asking other forum users if the puppy looked underweight. Responses discussed how to identify a healthy weight dog from appearances and the high prevalence of obesity distorting what is considered normal.

34
Mumsnet F3 An owner posted asking for weight reduction advice following an overweight diagnosis from their veterinarian. There was then a discussion with the owner about tailoring weight reduction advice to their life and how to recognize if their dog was a healthy weight.

Mumsnet F4
An individual posted about an acquaintance of theirs that had been underfeeding their dog accidentally. A discussion followed about how to decide how much to feed and confusion with following food package guidelines.

Mumsnet F5
A discussion was held about the high prevalence of overweight pets and the possible causes.

88
Mumsnet F6 An owner requested weight reduction advice. The responses focused on reducing calories consumed and increasing exercise.

22
Mumsnet F7 An individual posted asking for advice on how to motivate a family member to reduce the weight of their foster dog. A discussion followed about the negative impact of obesity and using shocktactics to encourage change.

Reddit F8
A poster sought advice and social support during their dog's weight reduction from other owners also attempting weight reduction.

16
Mumsnet F9 An owner asked whether their dog could be kept at a healthy weight whilst giving treats daily. A discussion followed about whether dogs can receive treats and be healthy.

Mumsnet F10
A discussion was held about the impact of excess weight on a dog's physical and mental wellbeing.

Mumsnet F11
A discussion about whether owners, or other factors, cause obesity in their pet dogs.

66
Petforums F12 An owner requested weight reduction and muscle building advice for an elderly dog. The discussion focused on the importance of exercise and nutrition for weight reduction and muscle gain.

Reddit F13
An owner asked for weight reduction advice for their recently adopted dog with joint problems. A discussion followed about how to exercise dogs with existing joint issues and food restriction for weight reduction.

Petforums F14
An owner asked for advice on how to change their parent's feeding behavior toward the poster's dog. The responses discuss who should be responsible for the dog's care and using shock-tactics in an attempt to stop excessive feeding.

21
Reddit F15 A documentary-type interview with the fosterers of a dog (described by the presenter as a beagle) followed their weight reduction journey, including a diet and exercise regime, at the end of the video the dog had lost 30% of its starting weight.

1,131 YouTube V1
A documentary-type interview with the owners of a dog (described by the presenter as a Rottweiler) focused on the dog's weight, shape, mobility, and diet. The dog was 100% over the average weight for its breed and sex.

124
YouTube V2 (Continued) were established to produce a better understanding of owners' perceptions, attitudes, and behaviors regarding canine obesity and weight management.
All the data used were in the public domain and used in accordance with the platforms' regulations. The University of Liverpool's committee on research ethics confirmed that their approval was not required for this study because the data were publicly available. To protect participants' anonymity, direct quotes have not been presented in the results section; instead, quotes have been slightly modified from the original whilst keeping the sentiment, and identifying details such as locations or names have been changed. For example, "my parents-in-law are terrible, they are always giving Poppy pieces of cheese" could be re-written as "my partners' parents often give Luna bits of cheese, they're awful." Further, instead of providing links to the forum threads and videos, the contents of these are summarized in Table 1. For example, in one forum thread (F1), an owner posted a photo of their young dog and asked other forum users whether the dog looked overweight and how much the dog should weigh. The responses included how to identify whether a dog is overweight from its appearance, whether young dogs should have a different body condition to adult dogs, and advice on weight reduction.

Results and Discussion
Four key themes emerged that showed individuals' behaviors around weight management were impacted by a number of perceptions and attitudes (Figure 1), including (1) Balancing conflicting responsibilitiesindividuals appeared to balance their responsibility in providing their dog with happiness, health, and love, and differences in emphasis placed on these impacted feeding habits and weight management; (2) Need vs. greed individuals felt compelled to alleviate perceived hunger in their dog, which made sticking to reduced food diets difficult for some; (3) Minimizingindividuals varied in the extent to which they perceived excess body fat to be problematic, and language used to describe their dog's body changed when excess body fat was seen as an issue; (4) Controlindividuals' perceived control over their dog's body condition and food intake varied hugely, with some owners believing they had little to no control. A couple and their adult daughter that shared ownership of three dogs were featured on a TV show where they discussed various aspects of the dog's care with the presenter. They were told that they had been overfeeding one of their dogs (described by the owner in the video as a Dachshund) and were advised to reduce the dog's food intake. A disagreement ensued about who was responsible for the dog's care.
1,007 YouTube V4 The owners of a dog (described by the owner in the video as a collie) took part in a weight reduction program as part of a TV show. The dog lost 15% of its starting weight and reached the target set by the presenter.

Balancing Conflicting Responsibilities
Whilst responsible ownership is generally understood to mean doing the "best for the dog," how this is achieved varies depending on individual owners' attitudes . We identified three key aspects of responsible ownership centered on the needs of the dog: ensuring health, happiness, and providing love. However, there were conflicting responsibilities around health, happiness, and love when it came to how owners fed their dogs. Some appeared to believe health and happiness were synonymous and, therefore, keeping a dog at a healthy weight was a way to ensure its happiness as well as their role as a responsible owner: Keeping your pet healthy and happy is surely responsible ownership and means your pet will live longer. F6 It was generally perceived that dogs enjoy eating and, therefore, many people appeared to believe that giving their dog extra food would increase its happiness. Some owners also used food to express love: Bandit was noticeably less happy after I moved away for Uni, my parents gave him lots of treats and scraps to cheer him up. F15 However, others believed that showing love should mean keeping a dog healthy. Several individuals discussed the dichotomy of expressing love through excess food, which could mean that the dog died young, and discussed that loving their dog meant they wanted them to live for as long as possible:  For some individuals, loving the dog and ensuring their overall happiness was valued more highly than the dog's physical health: My father's dog is massively overweight. Its only source of enjoyment is curling up on the sofa with my father and a treat. I don't see an issue, the dog is loved, cared for and generally happy. F6 If individuals perceived that providing food both enhanced the dog's happiness and expressed their love, weight reduction strategies that involved reducing food intake were difficult: My pug got fat recently, he loves eating but we had to cut down his food lots, it was heartbreaking F2 There is evidence within human obesity that the relative degrees of importance placed on different life domains are optimally reallocated as circumstances change so that a perception of overall subjective wellbeing is maintained; this is known as the adaptive global utility model (Bradford & Dolan, 2010;Dolan & Kavetsos, 2013). Our findings show that this may occur in response to canine obesity, too; so as a dog gains weight, the importance placed on their physical health is reduced and other domains such as happiness are more valued.

Need vs. Greed
Owners discussed observing several behaviors that they perceived as indications that the dog was seeking food, including begging, stealing food, and looking for food. Some owners interpreted these behaviors as their dog being hungry, whilst others believed that their dog was being greedy (Table 2). Hunger was perceived as the dog legitimately needing more food and as an unpleasant feeling for the dog, which often led to giving treats, extra meals, or bigger meals: I am of the opinion that it's better to overfeed than underfeed! I would rather a bit of extra chub than a hungry dog! F5 In contrast, greed was not perceived as a legitimate need for more food and was often referred to as an act or pretense; greed was also a label given to the dog, as if it was a constant personality trait: Table 2. Quotes showing the same behaviors could be interpreted as a dog being hungry or greedy.

Hungry Greedy
Begging "If she's begging at her bowl after she's eaten then I know she's still hungry" F9 "The fatter of my dogs is so greedy, she acts like she's never fed and begs constantly" F9 Stealing food "He was stealing from the bins he was so hungry" F5 "He's an awful thief and exceptionally greedy" F2 Looking for food "If he's looking for something and goes to his bowl then he must be hungry" F4 "He's so greedy, always looking for more food" F3 He's so greedy, always acting as if he's starving. He would keep eating and eating if he could and it's not because he's actually hungry. F2 During weight management, owners often believed that their dog felt hungry. In these cases, owners would often feel compelled to alleviate the dog's unpleasant feeling of hunger: I've worked out how many grams he needs and I'm sticking to that, he is not getting treats or snacks. The only problem is that he's hungry now. Is there anything else I can do to make him full? F4 Owners perceived that increasing satiety could be achieved through increasing dietary fiber, usually by adding vegetables to the dog's meal; less commonly by changing foods to a satiety-formulated diet: Cut down his main meal and add some veg. I use cooked green beans, they're great for increasing bulk and stop the dog feeling hungry. F7 Another option often used was enrichment feeding; many owners believed that dogs were both happier and more satiated the longer meals took to complete: What's worked for my dog is making her work for her food, her meals last longer and she always seems more satisfied than when she gobbles it in seconds. F4 Owners who did not give any extra food to a dog they perceived as hungry often shared concerns that restricting food may damage their relationship with the dog. They also mentioned that negative emotions were elicited, including guilt, sadness, and feeling mean or cruel: I know exactly what you mean about those pleading eyes when you've reduced their portions. I feel like I'm being really mean and I think the dog hates me for it. F14 When describing begging behavior, owners often referenced their dog's facial expressions, particularly their "puppy dog" or "pleading" eyes: I am guilty of giving treats. It's the way they look at you with those pleading eyes, it's hard not to give in. F8 A recent study suggests that "puppy dog eyes" are associated with eye movements rather than deliberate communication (Bremhorst et al., 2021). Thus, some researchers suggest that whilst the expression resembles sadness, it is not indicative of this emotion being experienced by the dog (Kaminski et al., 2019). In contrast, we find that owners perceive their dog to be communicating, and that trying to ignore these expressions during food restriction can elicit negative emotions in owners.
Whether owners perceived their dog as feeling hungry, or being greedy, led to some differences in how they responded and felt. Owners who thought their dog was greedy expressed less emotional turmoil over ignoring its apparent desire for more food. Several owners who perceived their dog to be greedy adjusted the environmental management to ensure the dog could not access extra food or did not have opportunities to beg: Our dog is always happy to eat more, she gets shut out of the room if she begs at the table V5 Although a difference in perception of hunger versus greed has not been observed in previous research, owners with negative emotional responses to perceived hunger in their dog during weight reduction may respond by feeding them more (Vedrine et al., 2021). Regarding interventions to prevent excessive feeding, changing owners' emotional responses to perceived hunger could be challenging given that compassion is a reflexive response to perceived suffering, and taking action to alleviate the suffering increases positive affect for the actor (Stellar et al., 2015). Therefore, instead of changing reactions to perceived hunger, finding ways to reduce signs of hunger could be effective and easier to achieve. One of the ways therapeutic weight-reduction diets promoting satiety are hypothesized to work is by reducing food-seeking behavior in dogs, thereby reducing owners' perception of hunger (German, 2016). In the current study, many perceived that enrichment feeding and making meals last longer also reduced their dog's perceived hunger between meals. Another approach would be to educate owners on how to minimize food-seeking behaviors.
The "present bias," whereby people tend to avoid immediate small costs even when the consequence is a greater future cost (Dolan & Kavetsos, 2013), might explain observations in the current study of tension in owners when dealing with the short-term cost of perceived hunger in order to achieve the longer-term benefit of successful weight reduction. In such circumstances, many owners appeared to opt for easing their dog's perceived hunger in the short term by increasing food intake, despite this potentially detracting from the dog's health in the longer-term. Again, reducing perceived hunger would address this issue.

Minimizing
Owners' understanding of their dog's body condition was complex, consistent with previous research in childhood obesity (Lachal et al., 2013) and equine obesity (Furtado et al., 2021). Many owners recognized that obesity was a risk factor for other health and wellbeing concerns. The most mentioned negative impacts were joint issues, reduced life expectancy, and reduced quality of life. These concerns provided motivation for some owners to both prevent dogs from becoming overweight and continue weight reduction for dogs that were overweight: I am really worried about his health, I know that being overweight can have a detrimental effect and if he stays this weight he might not live much longer. F2 However, individuals rarely discussed excess weight as an issue in itself. Instead, it was perceived to be a problem once other physical changes were noted that were impacting the dog's wellbeing, including struggling to breathe, having to stand or lie because sitting was not possible, and not being able to run or jump: Poor thing, he's not happy not being able to run and play. He tries to jump up when I come home but he's so fat he can't anymore. F6 When discussing how owners became aware that their dog was overweight, comments from a veterinarian or dog walker were most often mentioned: I was surprised when my vet said he was overweight, I hated the idea of him being overweight and tried to argue that he was muscular, not fat. But then I cut down his food and he got slimmer. F5 Some individuals used colloquial language that downplayed both the severity and significance of a dog being overweight. In this context, words used to describe dogs included "chubby," "chunky," "cuddly," "plump," and "porky." Further, whilst the use of such words suggested that owners were aware their dog was above their ideal weight, they appeared to see these descriptions as distinct from being overweight and sometimes explicitly stated that their dog was not overweight: Meg has always been more chunky than slight, not overweight exactly. F10 Veterinarians sometimes use humor and euphemism to discuss pet obesity, providing a means of facilitating conversations with owners about a potentially sensitive subject (Phillips et al., 2017). Such an approach might be counter-productive if owners interpret use of such terms to indicate that their dog is not problematically overweight. In the current study, there was divergence in the terminology owners used once they were at the point of actively attempting weight management: more clinical words such as "fat," "overweight," and "obese" appeared and rarely colloquial phrases that downplayed the severity of excess body fat. This, perhaps, suggests that owners shifted to using clinical terminology once they perceived excess body fat to require action. In human obesity research, the terms clinicians use when discussing obesity impact the patient's perceived self-efficacy to change, with a diagnosis of "obesity" leading to greatest self-efficacy (Hopkins & Bennett, 2018). Self-efficacy, in turn, is associated with successful weight reduction and maintenance in humans (Fontaine & Cheskin, 1997;Mitchell & Stuart, 1984). However, such terms can potentially be stigmatizing and problematic for patients, who negatively perceive terms such as "fatness," "obesity," and "excess fat" and prefer the use of "weight problem," "excess weight," or discussions about body mass index (Albury et al., 2020;Volger et al., 2012). Although similar research on how language impacts animal owner's self-efficacy is lacking, some authors have advised veterinarians to be careful about how they discuss obesity .

Control
Self-efficacy is an individual's belief in their own ability to exert control over their motivation, behavior, and environment in order to achieve desired outcomes (Bandura, 1977;Bandura, 1997), and this is an aspect of many behavior-change models (Kwasnicka et al., 2016). Greater self-efficacy is associated with successful weight reduction and maintenance in humans (Fontaine & Cheskin, 1997;Mitchell & Stuart, 1984). In the current study, there was considerable discussion around owner's perceived control over their dog's body condition and food intake. The extent to which individuals believed that a dog's food intake and body condition was within the owner's control varied: 1: No matter how much we cut his food he won't lose weight. We've tried but nothing works.
2: Sorry 1 but that's obviously drivel. Of course he'll lose weight if he ate less, you must still be overfeeding. F6 Some owners highlighted factors that might impact weight management that they had little control over, including breed, age, and health conditions, suggesting these individuals had low self-efficacy over their ability to achieve their desired goal of weight reduction: We have tried taking him for runs and put him on a diet. He's not fat because I'm abusing him, he's just a fat lab and it won't shift. F12 Many owners responded to suggestions that weight reduction could be achieved by reducing food by saying "I hardly feed him anything," whilst one participant, who identified themselves as a veterinarian, also stipulated that this was a common response to giving an overweight diagnosis. Such responses suggest either that some owners do not believe their actions have caused their dog to become overweight or that they have low selfefficacy over their ability to achieve the desired outcome by adjusting their feeding behavior.
Other owners acknowledged that their dog ate excessively but felt they had no control over this, often talking about the dog "sourcing their own food." For one owner, talking about this with their veterinarian increased their perceived control and empowered them to engage actively in weight reduction: I started her on a diet, my light-bulb moment was at the vets when I said "she eats too much" and he replied "you feed her too much." F6 Many owners also perceived they had little control over other household members feeding the dog treats, scraps, or extra meals: My mum acts like I'm starving him and she's always giving him treats and junk food when she thinks I'm not looking. V4 Individuals also discussed the dog's agency and suggested that excess weight was caused by the dog and that it should therefore take some responsibility for its actions: I'm fed up with this, the dog is an imbecile, he ate too much and now he's obese, it was his choice. V2 Weight reduction and management were often referred to as an ongoing struggle, with unsuccessful attempts eliciting feelings of futility, and advice-seeking in desperation, suggesting low perceived self-efficacy in individual's ability to successfully manage their dog's weight: My dog is really overweight, I don't know what I'm doing wrong. She doesn't eat much and I try to get her excited about exercise but she's not interested. I'm at my wits end. V5 Similar findings have been shown regarding pediatric obesity, with parents often turning to experts and weight-management programs only after failed attempts to reduce their child's weight or change their eating behaviors without external support (Davidson & Vidgen, 2017). Empowering children with obesity and their parents through collaborative goal-setting with their physician has been suggested for the successful treatment of pediatric obesity (Davis et al., 2007); this might be a suitable strategy for dog owners. The finding that some owners do not believe they are in control of their dog's food intake is a useful starting point, as reducing food intake is a critical aspect of weight management (Chapman et al., 2019). Such owners might require more detailed conversations and tailored advice on the actions they can take to reduce food intake.
There were strong attitudes and considerable stigmatization around body condition, with the target of stigma depending on whether individuals believed that the dog or the owner was in control. Stigma toward owners manifested as discussions about responsible ownership, often related to factors discussed above about ensuring health, happiness, and love for the dog. If owners were not perceived to be being responsible in these ways, they were often stigmatized and their suitability to care for a dog brought into question: RSPCA needed! Its abuse to overfeed a dog until it can hardly walk. This isn't love, she's obviously suffering. What a cruel and selfish owner. V2 Stigmatization toward humans with obesity has severe impacts on mental and physical health (Puhl, 2010). Such stigma is also associated with increases in food consumption, decreased perceived dietary control, and poorer outcomes in weight-reduction treatments (Major et al., 2014;Wott & Carels, 2010). Weight stigma by healthcare professionals is associated with shorter consultations, less information imparted, and patients avoiding preventive care due to feeling judged, disrespected, or embarrassed (Puhl, 2010). Weight stigma amongst veterinary professionals is associated with feelings of frustration and disgust toward owners and their dogs, along with a pessimistic outlook on owner compliance with weight-reduction advice (Pearl et al., 2020). Furthermore, media portrayals of canine obesity place the responsibility for treating obesity on the owners; this could increase stigmatization from the general public and veterinarians toward owners of dogs with obesity (Degeling et al., 2011;Degeling & Rock, 2012). It is possible that stigmatization of owners of dogs with obesity could result in similar negative outcomes to those seen in people with obesity. In this respect, classifying human obesity as a disease is associated with less weight stigma by healthcare workers and the general public (MacInnis et al., 2020;Nutter et al., 2018), possibly because highlighting biological causes of obesity, such as genetic factors, lessens the feelings of blame experienced by the patient Pegram et al., 2021). Thus, it is important that dog owners do not feel blamed or stigmatized for their dog's weight if they are to seek support and engage with advice. However, given that recognizing their role in controlling food intake is critical to weight reduction, this can be challenging, and expertise and resources in best practice to address this are required.

Future Directions, Strengths, and Limitations
This explorative study has revealed key areas where further qualitative research would be beneficial. Firstly, it is important to better understand how owners construct their role in caring for their dog and why they place emphasis on different aspects of their dog's care. Such research might facilitate the development of targeted intervention tools for owners with different perceptions. For instance, if owners believe that giving their dog treats is important for ensuring both their dog's happiness and their own status as a responsible owner, asking them to cut out all treats may be unachievable; instead, greater compliance might be achieved by promoting healthier treat options. Further research is also needed to explore how the perceptions of hunger and greed may be effectively addressed within a weight-management intervention. Understanding why owners attribute the emotion of hunger to their dog's behavior may enable the development of interventions that aim to address excessive feeding in response to perceived hunger. Furthermore, interventions should take the present bias into consideration and may need to focus on decreasing the perceived immediate costs of reducing food intake and increase salience of the benefits associated with weight reduction.
As approximately 50% of owners disagree with a veterinary diagnosis of overweight or obesity for their dog (White et al., 2011), further research is needed into how veterinarians communicate about obesity so as to engage owners better. Beyond recognizing whether their dog was overweight, there were also variable opinions as to whether excessive body fat was an issue. Further qualitative research is required to investigate how owners mentally construct issues around the consequences of obesity and to explore the motivations of owners that engage in weight management. From this knowledge, it might be feasible to develop interventions to improve owners' understanding of the consequences of excess body fat and motivate them to engage with weight reduction. Greater understanding of owners' perceptions about control and self-efficacy in managing their dog's weight will also be valuable for the development of an intervention tool that helps to increase owner empowerment and perceived control.
Analyzing online publicly available data has benefits, including ample availability of data online; minimal costs for data collection; the ability to observe the dynamic between participants in discussions; and, unlike other qualitative approaches, the presence of a researcher does not impact on the participants' responses. However, this lack of interaction between participant and researcher means there is no opportunity to ask for clarification; therefore, researchers may have to make greater assumptions about the meaning of comments made, and these interpretations might reflect beliefs of the researcher rather than the participant. To avoid this, the analysis was carried out by a team rather than a single researcher, and multiple data sources were used to ensure that themes were not limited to opinions from a single forum. The use of multiple data sources also somewhat addressed potential concerns with self-selection bias as the demographics of users vary across the different fora.

Conclusions
This study identified complexities in owners' management of their dogs' weight. Recognition of excess body fat is not straightforward, and several owners appeared to believe their dog was carrying excess body fat but did not perceive this to be a health concern. Individual recognition and understanding of excess weight as a problem impacted behaviors around weight reduction and management. The restriction of food, whilst easy for some, caused conflict for owners who believed that providing food was a means of expressing love for their dog and increasing its happiness. This was exacerbated by negative perceptions about dogs experiencing hunger, resulting in owners increasing food intake to satiate their dog or looking for alternatives such as enrichment feeding. Finally, owners varied in their perceptions of control over their dog's food intake and body condition, with responsibility not only put on others who feed it but also the dog. Further qualitative research is needed to understand how these findings may be used within weight-reduction, management, and prevention strategies.

Disclosure Statement
Alex German is an employee of the University of Liverpool, but his position is funded by Royal Canin, part of Mars Petcare. He has received financial remuneration and gifts for providing educational material, speaking at conferences, and consultancy work. Neither Royal Canin nor Mars Petcare were involved in any aspect of the study, including funding, design, execution, data analysis, interpretation, and drafting the manuscript. Robert Christley is an employee of Dogs Trust. Imogen Lloyd's PhD stipend is funded through Dogs Trust Canine Welfare Grants. Dogs Trust was otherwise not involved in the design, execution, data analysis, or interpretation of the data. Carri Westgarth is a consultant for Forthglade Pet Food and has received financial renumeration from both them and Royal Canin for provision of educational material.

Funding
This work was funded by Dogs Trust Canine Welfare Grants. The funder only reviewed the draft manuscript to ensure it was of appropriate standard prior to submission. It played no role in the study design, data collection, or analysis.