Feasibility and acceptability of a theory-based online tool for reducing stress-induced eating

Stress-induced eating is associated with various health risks like obesity and cardiovascular disease, exacerbated by the overconsumption of unhealthy foods. This study sought to investigate replacement coping strategies for stress-induced eating that participants can seek to implement using behaviour change techniques like implementation intentions. The study adopted a feasibility and acceptability design, with 258 participants (88.37% female) aged 17 – 75 years old who self-reported stress-induced eating. Participants were asked to identify cues for their stress-induced eating and evaluate the acceptability of eight potential replacement coping strategies. After selecting their preferred strategy, participants formed implementation intentions, linking the strategy with their previously identified cues. There were six themes of cues for stress-induced eating as identified by participants, including a range of external and internal stressors. Themes regarding the acceptability of the replacement coping strategies were organised based on constructs from integrated social cognition theories. Participant responses reflected cognitive and affective attitudes, and control and normative beliefs behind engagement in coping behaviour; further, automatic and volitional processes were described by participants as playing a role in whether a coping strategy was deemed as useful. Plans formulated by participants commonly detailed specific situations and strategies to utilise, though few described start times or durations of their plan. Action planning was found to significantly increase following formation of implementation intentions, and participants ’ descriptions supported the feasibility and acceptability of utilising implementation intentions to adopt alternative coping strategies to stress-induced eating. Future research should conduct a randomised controlled trial to assess the efficacy of the implementation intentions intervention in promoting uptake of replacement coping strategies to reduce stress-induced eating.


Introduction
Stress-induced eating has been characterised as a maladaptive behavioural response to stress in which food quality and quantity are altered due to experiences of stress (Evers et al., 2010;Kaiser et al., 2022;O'Connor et al., 2015).There is a tendency to overindulge in highly palatable, energy-dense "comfort" foods during such responses, which likely reflects a desire to induce more positive emotional states (Gibson, 2012).Accordingly, there has been concern for the prevalence of stress-induced eating behaviour due to the associated health risks exacerbated by the overconsumption of unhealthy food, such as obesity and cardiovascular disease (O'Connor et al., 2015;Qi & Cui, 2019).Owen et al.'s (2020) study during the coronavirus disease 2019 (COVID-19) pandemic reported that 53.6% of Australians reported overeating or having poor appetite over a 2-week timeframe, attributing this to the impacts of lockdown, high anxiety, and fear of contracting COVID-19.Further, university students with high perceived stress during COVID-19 were commonly found to seek out highly palatable foods compared to those with lower perceived stress (Cheng & Wong, 2021).The concerning relationship between stress and unhealthy eating behaviours indicates a need to research and understand engagement in stress-induced eating.

Eating in response to stress and negative emotions
Various conceptualisations of eating unhealthy foods in response to stress or negative emotions exist in the literature.For example, emotional eating has been defined as eating in response to negative emotional experiences such as sadness, anger, or boredom (Evers et al., 2018).Conversely, binge eating has been defined as episodes of overeating which is accompanied by a sense of loss of control (Haedt-Matt & Keel, 2011).Escape theory suggests that reductions in negative affect occur during binge eating episodes by directing the focus of cognitive attention from high level abstract thinking to the immediate situation, with distress returning upon cessation (Heatherton & Baumeister, 1991).The affect regulation model takes this conceptualisation further, positing that engagement in binge eating behaviour occurs to alleviate negative affect by utilising food to provide comfort and/or distraction; and subsequent increases in negative affect following cessation of eating then serve to reinforce engagement in the behaviour (Hawkins & Clement, 1984).Conversely, stress-induced eating has been defined as changes in food quantity or quality in response to experiences of stress (Kaiser et al., 2022).For the purposes of the current study, we have defined the target behaviour as stress-induced eating whereby the eating occurs in response to a stressful experience, which may or may not result in the individual experiencing negative emotions they are aiming to attenuate with the behaviour.The rationale for this approach is that conceptualisations of emotional eating require that the eating occurs in response to an emotion, without the presence of a stressor being formally specified.

Theoretical foundations for understanding stress-induced eating
The relationship between stress and eating can be understood through the lens of the transactional model of stress and coping (Lazarus & Folkman, 1984).Lazarus and Folkman (1984) described a three-part process where individuals appraise the demands of a situation (primary appraisal), consider their available resources to cope (secondary appraisal), and then make efforts to cope with the situation.Such coping efforts are commonly classified into problem-and emotion-focused coping, centring on whether direct action is taken to deal with the stressor or to reduce the emotions arising from a stressor, respectively.Further, these responses can either be adaptive or maladaptive in nature (Carver et al., 1989;Lazarus & Folkman, 1984).Engagement in stress-induced eating has been suggested to stem from a lack of more adaptive coping strategies being available (Evers et al., 2010;Ozier et al., 2007), and so it follows that reducing this behaviour may be done by imparting healthier coping strategies for individuals to employ when stressed.However, Lazarus and Folkman's (1984) model did not specify the underlying mechanisms between the stress appraisal process and the coping response that may influence whether a given coping strategy is implemented during a stressful situation.Social cognition theories have widely been applied to understanding the determinants of health-related behaviour (Hagger et al., 2020), and can provide valuable insights into the proximal determinants of coping responses when applied in this context (Keech & Hamilton, 2022).
The theory of planned behaviour (Ajzen, 1991), a prototypical social cognition theory, proposes that engagement in a behaviour is predicted by an individual's intentions to perform the behaviour, and in turn, intentions are influenced by three psychological factors.Attitudes are the positive and negative appraisals regarding a behaviour, underpinned by behavioural beliefs (costs and benefits); and can be further categorised as cognitive (thoughts/ideas/perceptions) or affective (emotions/feelings).Subjective norms comprise of the social influence of significant others, underpinned by normative beliefs (the approval/disapproval of others).Finally, perceived behavioural control encompasses the perceived ease or difficulty of performing a behaviour, underpinned by control beliefs (barriers and facilitators).These underlying beliefs behind intention and behaviour provide targets which can be intervened upon in behaviour change interventions (Hamilton et al., 2023).Previous research has applied the theory of planned behaviour to predict a range of health behaviours including food-related behaviours (McKee et al., 2019;Mullan et al., 2014;Spinks & Hamilton, 2015) and as such, it can serve as a useful theoretical framework to understand the reasoned processes behind the uptake of replacement coping strategies for stress-induced eating.
However, given that intention is not always translated into behaviour, recent research has also suggested a need to look beyond these reasoned, intentional processes to consider more automatic, nonconscious determinants of behaviour (Hagger & Chatzisarantis, 2014;Keech & Hamilton, 2022;Mullan et al., 2014;Rackemann, Hamilton, & Keech, 2024).Dual-process theories suggest that two systems of cognitive processing exist-Type 1 and Type 2 processes-which refer to automatic and reasoned processes, respectively (Evans & Stanovich, 2013;Strack & Deutsch, 2014).In times of stress, where working memory capacity is under high load, behaviour is thought to be regulated by the more automatic, rapid Type 1 processes that requires lower levels of cognitive capacity than Type 2 processing (Banks & Boals, 2017;Evans & Stanovich, 2013).Addressing the automatic processes underpinning behaviour-through constructs like behavioural automaticity and habit formation-may then be particularly vital when designing interventions for changing coping behaviour (Keech & Hamilton, 2022).
Volitional processes such as planning can also help to bridge the intention-behaviour gap by promoting recall and enactment of intention to engage in a behaviour (Hagger & Chatzisarantis, 2014).Planning strategies such as implementation intentions have been identified as a potentially valuable strategy for targeting the reasoned and automatic processes behind behaviour (Keech & Hamilton, 2022;Rackemann, Hamilton, & Keech, 2024).Implementation intentions assist individuals to monitor cues and create "if-then" behaviour plans, serving to promote awareness and in turn, may encourage habit formation and behavioural automaticity through continued repetition of a plan (Gardner et al., 2020;Gollwitzer & Sheeran, 2006;Keech & Hamilton, 2022;Rackemann, Hamilton, & Keech, 2024).Previous studies have found implementation intentions to be successful in replacing a range of habitual behaviours, such as recycling and unhealthy snacking (Adriaanse et al., 2009;Holland et al., 2006).This suggests that there may be merit in utilising implementation intentions to reduce stress-induced eating, accounting for automatic, reasoned and volitional processes underpinning behaviour.

Approaches to changing eating behaviour
Previous research has investigated the effectiveness of implementation intentions for replacing undesirable eating behaviours with more adaptive behaviours (Adriaanse et al., 2009).Adriaanse et al. (2009) asked participants to identify motivational and situational cues behind their unhealthy snacking behaviour, and then form implementation intentions to reduce their unhealthy snacking in favour of healthy snacking.Results indicated that unhealthy snacking was effectively reduced when individuals formed implementation intentions specific to their personal motivational cues, rather than being assigned a common cue identified across participants (Adriaanse et al., 2009).Gollwitzer (1999) explained that the reasoned "if-then" process of linking cues with a desired behaviour promotes automaticity by facilitating the formation of a "mental link".In considering Adriaanse et al. (2009) findings, it may be that a "mental link" is likely to be stronger when implementation intentions are congruent with an individuals' personal cues, thus promoting effective behaviour change.
Implementation intentions have also been investigated in relation to replacing emotional eating behaviour.Armitage (2015) found that in forming implementation intentions linking triggers for emotional eating with more appropriate behavioural responses, participants' emotional eating in response to boredom was significantly reduced.Similarly, one study has examined the effectiveness of implementation intentions to encourage healthy snacking as a replacement for unhealthy stress-induced eating (O'Connor et al., 2015).O'Connor et al. (2015) had participants form "if-then" plans by identifying cues for their stress-induced eating and a healthy snack alternative to eat in response to stress.Overall, participants' unhealthy snacking in response to daily stressors was reduced; however, increased healthy snacking in response to stress was more strongly observed in individuals motivated towards healthy eating (O'Connor et al., 2015).These findings highlighted the importance of considering individual preferences towards coping strategies when prompting individuals to form an implementation intention in this context.Therefore, it may be beneficial to provide a broad range of replacement coping strategies that individuals may select, in order to accommodate individual preferences.

Potential replacement coping strategies for stress-induced eating
In reducing unhealthy eating in response to stress, previous studies have independently investigated the utility of healthy eating (consuming nutritious foods like fruit and vegetables) and physical activity (any movement of skeletal muscles that requires energy expenditure) (Horsch et al., 2015;Leow et al., 2018;O'Connor et al., 2015;World Health Organisation, 2022).Additionally, studies have also evaluated mindfulness and relaxation techniques (non-judgemental acceptance and attention given to the present moment) for stress-induced and emotional eating behaviour (Corsica et al., 2014;Katterman et al., 2014).Research has also highlighted the utility of emotion regulation strategies in reducing negative affect and stress (Evers et al., 2010;Van Bockstaele et al., 2020).An example of such strategies include reappraisal, in which thought processes regarding a situation are changed (Evers et al., 2010).Situation modification has also been investigated, which in contrast to reappraisal is a problem-focused coping strategy where the focus is instead on modifying the physical aspects of a situation (Van Bockstaele et al., 2020).
To extend beyond the limited scope of existing strategies examined within the stress-induced eating literature, the current study derived three potential coping strategies from the Brief-Coping Orientation to Problems Experienced inventory (COPE; Carver, 1997), identified for their ability to reduce stress and thus, potentially reduce stress-induced eating indirectly.Additionally, it was required that these strategies were not health detrimental.This included: humour, defined as the perception or expression (i.e., laughter, making jokes) of the humorous features of a stressor (Canestrari et al., 2021;Simione & Gnagnarella, 2023); instrumental (seeking practical help/advice) and emotional (seeking comfort/understanding) support (García et al., 2018;Mathieu et al., 2019;Shrout et al., 2006); and positive distraction, defined as distracting oneself with activities that induce positive emotion so as to be better able to manage a stressor later (Waugh et al., 2020).This study sought to investigate the acceptability of replacement coping strategies that were not unhealthy and potentially effective in reducing stress-induced eating to inform future intervention development.

The current study
Prior research has supported the utility of implementation intentions in reducing unhealthy snacking in general (Adriaanse et al., 2009).A further study demonstrated the utility of implementation intentions in reducing unhealthy snacking in response to stress specifically (O'Conner et al., 2015).O 'Connor's et al. (2015) study also observed greater uptake of healthy snacking as an alternative coping response among individuals who were motivated to eat healthy.However, these studies also highlighted the potential benefit of participants self-selected cues in such implementation intentions plans (Adriaanse et al., 2009), and that individuals may differ in their preferences to adopt particular replacement coping strategies (O'Conner et al., 2015).Therefore, the online intervention tool designed for the current study prompted participants to self-select their cues for stress-induced eating, and provided a range of potential replacement coping strategies to cater for individual differences in coping strategy preference.The aim of this study was to examine the acceptability and feasibility of an online tool designed to support individuals to replace stress-induced eating with an alternative coping behaviour.First, given the importance of cues in habit formation and planning (i.e., implementation intentions), we sought to examine the cues that individuals identify as prompting their stress-induced eating (Research Question (RQ) 1).Second, we investigated the acceptability of various coping strategies that individuals may use to replace stress-induced eating, and the utility of these strategies when they are paired with implementation intentions (RQ2).Third, we sought to evaluate the effect of an implementation intentions exercise on the formation of a specific plan for enactment of the chosen coping behaviour, as indicated by a measure of action planning (RQ3).It was expected that action planning would significantly increase following completion of the implementation intentions exercise.Fourth, the plan quality of the implementation intentions formulated by participants was also analysed to provide an understanding of how well the planning exercise was completed (RQ4).Finally, open-ended feedback from participants on their experience using the tool was analysed (RQ5).

Participants
The sample (N = 258) consisted of individuals aged 17-75 years old (M = 30.92,SD = 13.93)(88.37% female) who self-reported engaging in stress-induced eating.The sample was predominantly Australian (93.3%).Self-reported body-mass index (BMI) scores ranged from 15.62 to 51.86 (M = 26.63,SD = 7.04).Five participants were not included in the BMI calculation due to providing weight and height data combinations which were deemed as likely to be erroneous.Recruitment occurred via social media platforms such as Facebook and Instagram, and student and staff newsletters at a major university in South-East Queensland, Australia.The majority of participants (61.48%) were also recruited through a subject pool where first-year psychology students were able to receive course credit upon participation in research projects.In terms of sample size, we aimed to collect as much data as possible with the resources available and in the available time period (4 months).There were two further considerations in ensuring sample size was adequate.The first of which was for the qualitative data.While there is considerable debate in the published literature on how to determine sample size for qualitative research, we followed guidelines by Braun andClarke (2013), andFugard andPotts (2015), who recommended a minimum of 10-50 participants for participant generated text.We sought to collect enough data to ensure at least 10 participants had selected each of the eight replacement coping strategies (the least frequently selected coping strategy was selected by 13 participants).The second consideration was adequate power to conduct the t-tests for preand post-intervention action planning.An a priori power analysis using G*Power v3.1 suggested that 156 participants would be required to detect a small effect size (d = 0.20; Power = 0.80; two-tailed repeated measures t-test).Therefore, when 258 participants were recruited in the planned recruitment period, it was determined that no further data collection was required.Due to recommendations by Braun and Clarke (2021) for reflexive thematic analysis, we did not use data saturation as a determinant of sample size.Namely, Braun and Clarke (2021) identified that there is lack of agreement on how to evidence what constitutes information redundancy and that the concept of saturation is not aligned to the paradigm underpinning reflexive thematic analysis.Rather, they recommended to focus on information power (i.e., collecting more data for shorter text responses) and pragmatics, which underpinned our sample size determination.

Design and procedure
The study was designed to determine feasibility and acceptability of an online intervention tool.Participants completed a 30-min online, anonymous survey through Qualtrics online survey software.Participants began the survey by self-reporting their stress-induced eating over the last week, rating the occurrence of behaviour on a scale of 1 (never) to 7 (always).Stress-induced unhealthy eating was defined as "when you eat unhealthy snacks (such as chocolate, potato chips, cakes, ice cream, lollies) due to experiencing a stressful situation.This can be any type of stressful situation (e.g., an approaching deadline at work; an assignment needing to be completed; a disagreement with a friend)."Participants who responded 1 (never) on the scale were ineligible and were redirected out of the survey (those who responded seldom, rarely, sometimes, often, very often, or always, were included).This was to ensure participants were currently performing the target behaviour and that the intervention would be relevant to them.Those not meeting the minimum age requirements (i.e., aged 17 years and older) were also excluded.The study was designed to sample adult participants.However, we opted to include 17-year-olds in the study due to a large number of 17-year-olds in the region where the study was conducted either being enrolled in first year university studies or entering the workforce at 17 years old, having finished high school at that age.We therefore determined that the experience of 17-year-olds should be similar to that of 18-year-olds.Ethical guidelines also provided that those aged 17 years old were able to provide informed consent to participate themselves, rather than requiring parental consent, as would be required for younger participants.
Part 1 of the online tool had participants identify five cues behind their stress-induced eating behaviour.Part 2 involved providing brief descriptions of eight replacement coping strategies, and participants were then prompted to evaluate and reflect upon the utility of each strategy.Part 3 then asked participants to select their preferred coping strategy and formulate an implementation intention plan linking their chosen strategy with their previously identified cues for stress-induced eating.Participants then provided feedback on the exercise, with prompts to comment on the utility and difficulty of the exercise, and to share their suggestions for improvements.At each stage, qualitative content analysis was conducted.The effect of the implementation intentions exercise on action planning from pre-post completing the exercise was assessed quantitatively.The quality of the plans formed during the implementation intentions exercise was also assessed.Table 1 provides further detail on the development of the online tool.This study was not preregistered, though the data file and study materials were made available on the Open Science Framework (https://osf.io/qvnbh).

Qualitative responses
In working through the online intervention tool, participants provided qualitative open-ended responses to questions about cues for stress-induced eating and their evaluation of each of the eight replacement coping strategies.Participants then formulated and recorded an implementation intention plan to implement their chosen strategy to replace stress-induced eating.They were then asked to answer five openended questions addressing difficulty and utility of the exercise, and suggestions for improving the task.

Action planning
Participants self-reported their action planning regarding when, where, how, and how often they planned to use their chosen strategy (e. g., "I have made a plan regarding … when to use this strategy") (Schwarzer, 2008).There were four items rated on a 7-point Likert scale, with scores ranging from 1 (not at all true) to 7 (exactly true); higher scores indicated higher planning.Total scores were calculated by computing the individual's mean score across the four items at each times point.The scale demonstrated adequate reliability at both time points (T1 α = 0.94; T2 α = 0.94).While action planning and implementation intentions have some conceptual differences, scholars have argued that action planning as a measured construct has considerable overlap with implementation intentions (Hagger et al., 2016).Schwarzer (2008) and Schwarzer et al. (2008) have further demonstrated this overlap by referring to implementation intentions focused on initiation of action as action plans.Given that no specific measure of implementation intentions exists, Schwarzer's (2008) measure of action planning was selected for use as a manipulation check regarding the extent to which participants perceived themselves as having a specific plan for using their chosen strategy.et al., 2018;Mathieu et al., 2019;Shrout et al., 2006); ( 7) humour (Belt, 2006;Canestrari et al., 2021;Simione & Gnagnarella, 2023); and (8) distraction ( Carver & Scheier, 1994;Waugh et al., 2020).

Demographic information of participants was collected regarding
Strategies were selected based on prior research examining their utility in reducing stress and/or changing eating behaviour.The aim was to provide a comprehensive list of replacement coping strategies, inclusive of both problemand emotion-focused coping strategies (in accordance with Carver's (1997) Brief-COPE), that were not health-detrimental and were potentially adaptive.

3: Planning
Implementation intentions exerciseindividuals created action plan linking their chosen replacement coping strategy with previously identified cues for stress-induced eating Based on previous research on implementation intentions exercises (Gollwitzer, 1999;Gollwitzer & Sheeran, 2006).Implementation intentions have been utilised for changing a broad range of behaviours (Rackemann, Hamilton, & Keech, 2024), but also specifically in studies aiming to change eating behaviour (i.e., Adriaanse et al., 2009).Quality of plans formed during implementation intentions exercise was assessed, using an adapted coping scheme from Rackemann, age, gender, country of residence, and body mass index (BMI).

Data analysis
Data were imported and prepared using SPSS V28.Twenty-seven participants were screened out due to not meeting eligibility criteria.Eight participants were removed due to indicating that they did not meet eligibility criteria for stress-induced eating in an open-ended response.An additional twelve participants were removed due to incomplete open-ended responses.The survey also included two attention check questions.Thirty-four participants who did not pass either of the attention checks were removed.
Qualitative data were analysed using thematic analysis.An inductive/deductive approach was taken.This involved first taking a datadriven approach to inductively identifying themes, and then organising the categories of themes deductively based on an integrated social cognition approach (Hagger et al., 2020).We followed Braun and Clarke's (2006) recommended six stages of qualitative analysis: (1) familiarise yourself with the data, (2) generate initial codes, (3) search for themes, (4) review themes, (5) define and name themes, and ( 6) produce the report.The analysis was guided by this approach, with responses coded using NVivo (Version 12) and initially sorted into preliminary themes.Following the iterative process of refining and editing themes, categories were then created to encompass broader patterns of the themes.These categories were organised according to the integrated social cognition approach.Quantitative data for action planning were analysed using a paired samples t-test analysis in SPSS V28.Frequencies for participant coping strategy preference were also produced using SPSS.Implementation intention plan quality was then coded using NVivo according to a coding scheme adapted from Rackemann, Hamilton, & Keech, 2024.

Research Question 1: cues for stress-induced eating
Six broad themes were derived from participant descriptions of their cues for engaging in stress-induced eating: (1) specific tasks to complete, (2) interpersonal situations, (3) internal cues, (4) daily tasks/hassles and stressors, (5) health-related concerns, and (6) desire to seek positive affective states.These themes encompassed multiple subthemes, which help to highlight their contextual nuances, and are outlined Table 2. Bolded text indicates sub-theme titles.

Specific tasks to complete
This theme incorporated cues in which stress arose from the perceived external pressures to meet deadlines and complete tasks.Commonly, this was in relation to university and/or work.Stressinduced eating in such situations was described by participants to be an automatic, learned pattern of behaviour.Participants also commonly reported that unhealthy food was quicker and more accessible compared to healthier snacks and as such, time pressures resulted in little attention given to preparing nutritious snacks/meals.

Interpersonal situations
This theme was defined by participants' descriptions of stressful interactions with others such as family, friends, and work colleagues.Participants referenced stress caused by both direct and indirect interactions with others, which prompted eating to escape or reduce negative emotional experiences.Disagreements and tension with others were a commonly identified stressor.Some participants also identified that isolation from others motivated their unhealthy eating, an indirect stressor that manifested from feelings like loneliness or social inadequacy.Social events were also identified as a stressor, where stressinduced eating occurred in anticipation of or immediately following socialising with others, commonly reflecting feelings of social anxiety or pressure.

Internal cues
The theme of internal cues encompassed stressors of uncertainty, low energy (physical and/or mental), concerns around dieting and health (including a sub-theme around not eating sufficiently), negative self-image, and general negative affective states-some which involved the presence of symptoms of psychological disorders.These stressors were described by participants to be more internal, with these emotional experiences functioning as the cue for stress-induced eating themselves, rather than an external factor (as can be seen in other themes).The function of stress-induced eating in such situations reflected a desire for comfort, where unhealthy "comfort" foods provided hedonistic pleasure and were utilised to self-soothe negative affect.This response was commonly thought by participants to be automatic, reinforced by past associations made between stress and food.

Daily tasks/hassles and stressors
This theme incorporated daily life stressors.Responsibilities to maintain a home (domestic tasks) and family were a commonly identified stressor described as a cue for stress-induced eating.This encompassed carer responsibilities like putting children to bed or driving them to school, extracurricular activities and concern for their wellbeing.Financial concerns were also a common stressor.

Health-related concerns
This theme reflects sub-themes comprising the health and wellbeing of oneself and others (i.e., family members).Stress associated with such events reflected concerns over health but also, lowered motivation to eat healthy and resist the temptation to engage in stressinduced eating behaviour for comfort.

Desire to seek positive affective states
This theme encompassed the desire to achieve more positive emotional states, where unhealthy snacking functioned as a reward and/ or motivation to complete tasks.Whilst this was a general sub-theme that emerged within each theme, this theme in particular reflected Note.N = 258, percentages (%) = n/N.
T.L. Maroney and J.J. Keech patterns of stress-induced eating where hedonistic pleasure offered by unhealthy snacking was relied upon to reduce stress and negative emotional states arising from stressors.

Research Question 2: factors underpinning acceptability of alternative coping strategies
As outlined earlier, the themes in the data on acceptability of alternative coping strategies were organised based on an integrated social cognition theoretical approach which integrates beliefs from the theory of planned behaviour (behavioural, normative, and control beliefs), and automatic (i.e., habit) and volitional processes (i.e., planning).Tables 3-5 provide a complete outline of all themes and sub-themes derived from the data, organised based on the theoretical constructs above.The most salient themes, which were present across at least four of the eight coping strategies, are described in detail below.However, it is notable that sub-themes within the normative beliefs and volitional processes themes were present in less than half of the eight coping strategies, but were still reported to provide a comprehensive understanding of their roles in the acceptability of a replacement coping strategy.Exemplar quotes for each theme and sub-theme are presented in Supplementary Appendices C through F.

Behavioural beliefs
Affective Attitudes.This theme encompassed attitudes regarding the anticipated affective consequences (e.g., increased or decreased negative or positive affect) of using a given strategy.Across all strategies, participants reflected upon the potential for improved mood or increased positive affect, including ideas of enjoyment-for example, the taste of food in reference to the healthy eating strategy.For seven of the eight strategies, participants also described considering whether a strategy may actually cause more stress and/or negative affect (e.g., a lack of success with the replacement coping strategy may lead to further stress).A strategy's ability to reduce negative affect (e.g., guilt) was also considered by participants across half of the strategies.
Cognitive Attitudes.This theme encompassed beliefs about the advantages and disadvantages of the replacement coping strategies.Present for all strategies was the consideration of utility, where participants reflected generally upon whether they believed a strategy would/would not be useful or were uncertain/hesitant in this evaluation.Participants also considered whether a replacement coping strategy might only be a temporary solution.Across more than half of the strategies, participants reflected on how a strategy may function as a distraction or avoidance from their stressors or associated negative emotions.In some contexts, this was described favourably, whereas in others, this was a negative.The participants who described this as a negative, indicated that use of the strategy would not deal with their stressors (i.e., avoidance), which would cause more stress.Further, participants considered that they may find it difficult to return to the directly dealing with their stressor as they would continue avoiding it using the replacement strategy.There was also a sub-theme of participants believing that their eating behaviour would not be reduced when utilising these replacement coping strategies.In some cases, individuals thought that their eating may actually increase, such as in regard to using physical activity as a replacement coping strategy.For example, one participant described that hunger may result from energy exertion, or food may be utilised as a reward for completing exercise.Participants also reflected upon whether a strategy could change how they perceive a stressor and their ability to manage it (e.g., 'This would most likely be useful to me because breaking down a stressful situation logically may dispel some of the emotional catastrophising').

Control beliefs
Facilitators.This theme encompassed sub-themes reflecting the perceived facilitators to utilising a replacement coping strategy, such as successful past experiences and perceived applicability across situations.For all strategies, a key sub-theme was participants' reflection on past and/or current success with the coping strategies and how this may influence the likelihood that they would enjoy or find a strategy effective.
Barriers.This theme contained sub-themes pertaining to the perceived barriers (e.g., accessibility, motivation, stressor/strategy fit) to utilising a replacement coping strategy.Key sub-themes were related to accessibility (including a sub-theme of cost specific to healthy eating), ability to engage (such as motivation and energy) and the stressor itself (e.g., intensity, changeable versus unchangeable  stressors).Additionally, present in just over half of the strategies was a sub-theme regarding the perceived difficulty of a strategy; a sub-theme within perceived difficulty was the need for external support, in which participants believed they may need guidance from another individual or external factors (e.g., reminders) to engage with a strategy.Across half of the strategies were considerations negative past experiences with a strategy, where participants considered how such experiences influence their perceptions of a strategy's utility.Participants also considered the need for a particular headspace/mindset to engage with a strategy.

Normative beliefs
This theme comprised of one sub-theme reflecting the social influence of significant others, like friends and family.Such beliefs were only considered in reference to the seeking support strategy, where some participants reflected upon how they may feel they would be burdening others with their stress.

Automatic processes: reflections on habits
Automatic considerations included reflections of the automaticity of current coping behaviour.Across over half of the replacement coping strategies, participants considered the automatic nature of their eating, and how this habit/automaticity of initiating the behaviour without thinking makes it harder to exercise self-control.As such, participants expressed that they would need to be more aware of their behaviour in order to implement a new coping strategy in place of their stress-induced eating.Participants also considered that in times of stress, there was often a lack of self-control and so their eating behaviour was often outside of their awareness.As a result, participants thought that a given strategy would likely not be automatically utilised when stressed.

Volitional processes: action planning
This theme encompassed sub-themes related to the volitional processes behind behaviour, with participants reflecting on the need for action planning to encourage habit and avoid forgetting the new behaviour.Notably, the key sub-themes within this category were all present across just under half of the strategies.Sub-themes indicated that participants considered the need for planning ahead and the likelihood of forgetting to engage with the strategy, in turn reflecting a need for planning.For both the healthy eating and mindfulness strategies, participants commented on the need to make a conscious decision and effort to engage with the strategy, once again reflecting the importance of formulating a plan.

Preferred coping strategy
Fig. 1 provides a graphical representation of replacement coping strategy preferences among participants.Majority of participants indicated a preference for physical activity (32.9%) and distraction (22.2%), with the remaining six strategies each chosen by nine percent or less.

Research Question 3: action planning
A paired-samples t-test was also conducted to determine whether action planning was influenced by the implementation intention exercise.There was a significant difference between action planning scores across the two time points, t(227) = 5.06, p < 0.001, d = 0.34.Time point two (T2; M = 5.17) scores were 0.33 higher than time point one (T1; M = 4.84) scores (95% CI [0.20, 0.47]), indicating that participants were significantly influenced to formulate a plan to engage with their chosen strategy following formation of their implementation intentions.Based on Cohen's (1988) guidelines, this represents a small to medium effect size.This effect size is smaller than the large within subjects effect size (d = 0.83) for change in action planning observed by Rackemann, Hamilton, & Keech, 2024 pre-and post-prompting participants to create implementation intentions plans to increase their use of problem focused coping.

Research Question 4: implementation intention plan quality
The implementation intentions plans generated by participants were analysed to determine the quality of the plans made.Responses were coded according to a coding scheme that was adapted for the current context based on Rackemann, Hamilton, & Keech, 2024.This involved identifying whether a specific situation and strategy from the list provided were accurately described, as well as whether a start time and duration were specified (see Supplementary Appendix A for the coding scheme).The results indicated that 64.22% participants detailed a specific situation and 82.83% identified a specific strategy (see Supplementary Appendix B for further information on the frequencies and percentages of plan quality indicators).A smaller number of participants indicated a specific start time (7.73%) and duration (6.44%) in their plan.

Research Question 5: general feedback on using the online tool
Following completion of the online survey, participants were prompted to provide feedback regarding their thoughts on the exercise, perceived difficulty/ease of completing the task, whether they believe exercise will promote change in their behaviour within the next week, and any ideas for improving the task.These questions were intentionally kept broad to prevent leading the participants to provide positive feedback (i.e., demand characteristics) and to allow participants to freely report their most salient thoughts on the acceptability of the online tool, and potential areas for refinement.Percentages reported below represent the proportion of participants whose descriptions aligned to the respective theme or subtheme.Percentages are generally small as this was based on free description by participants, and every theme was not put to participants as a question.Four broad themes were derived from available participant feedback: (1) ease of completion, (2) implementation intentions, (3) likelihood of behaviour change, and (4) replacement coping strategies.

Ease of completion
This theme encompassed sub-themes regarding participants' perceived ease of completion of the online tool and general survey.Majority of participants did not commonly identify any difficulties with answering questions or completing the implementation intentions activity (81.01%), with some reflecting that it was simple and straightforward (8.91%).

Implementation intentions
This theme pertained specifically to the implementation intentions activity.Of participants who provided feedback explicitly on the exercise itself, 12.40% expressed that they enjoyed formulating their plan/ goal.

Likelihood of behaviour change
This theme encompassed participant responses to whether they believed their behaviour would be influenced by using the online tool.Thirty-one percent of participants considered that the online tool facilitated reflection, which increased their awareness and accountability for their behaviour.Sixteen percent of participants also expressed intention to try to utilise their chosen replacement coping strategy, with 12.40% specifically indicating they had a plan in place or expressed certainty in implementing their strategy.Ten percent of participants believed they were only moderately influenced, and 8.91% expressed that they were not influenced/their behaviour would not change.A small number of participants indicated they felt more confidence or had an increase in motivation to change (5.81%).

Replacement coping strategies
Sub-themes within this theme comprised participant reflections of the replacement coping strategies, particularly regarding their utility and feasibility in practice.Some participants considered the utility/ benefits of implementing multiple strategies (9.30%)-in which they reflected upon using different replacement coping strategies depending on the stressor-while 7.36% of participants reflected generally upon the benefits of their chosen strategies or the potential barriers/difficulties in applying their strategies when experiencing stress (5.43%).

Discussion
The central aim of this study was to investigate the acceptability and feasibility of an online tool in supporting the replacement of stressinduced eating with an alternative coping behaviour that is not health-detrimental.Qualitative analysis of the cues behind stressinduced eating provided valuable insight into the engagement and maintenance of the behaviour, which may be utilised in future interventions to assist individuals in formulating specific plans for changing coping behaviour.Regarding the acceptability of replacement coping strategies for stress-induced eating, qualitative analysis identified that participants considered several specific cognitive and affective attitudinal factors, normative factors and barriers and facilitators, for determining whether they would use a given coping strategy.Further, participants also considered the habitual and automatic nature of behaviour and relatedly, the need for planning, when reflecting upon a strategy's utility.Accounting for such factors may be valuable in future intervention development, allowing interventions to best encourage the uptake of replacement coping strategies when seeking to change behaviour.Self-reported action planning significantly increased from pre-to post-administration of the implementation intentions exercise, which provides preliminary support for the use of implementation intentions to target stress-induced eating.Analysis of implementation intention plan quality indicated that the majority of participants identified specific situations and coping strategies when formulating their plans, but few detailed specific start times and durations of the plan.This provides clear direction for refining participant instructions to support formulation of a detailed implementation intention plan in future research.

Cues for stress-induced eating
Existing research has found that implementation intentions formed around triggers and personally relevant motivational cues were able to reduce both unhealthy snacking and emotional eating behaviours (Adriaanse et al., 2009;Armitage, 2015).Interventions may then best be equipped to aid individuals in changing their behaviour by holding an understanding of the range of cues prompting engagement in stress-induced eating.This study identified six categories of cues as indicated by participants, with five categories encompassing external cues-specific tasks to complete, interpersonal situations, daily tasks/hassles and stressors, health-related concerns and desire to seek positive affective states-and one category reflecting internal cues which comprised emotional experiences functioning as the stressor.Regardless of the stressor, participants generally expressed that their eating response was a distraction or escape technique from their stress and associated emotions.This is largely consistent with previous research suggesting eating in response to stress is motivated by a desire to alleviate negative affect, leading to the consumption of highly palatable foods for comfort and more positive affect (Gibson, 2012;O'Connor et al., 2015).This finding also suggests the potential value of augmenting dual-process social cognition accounts of stress-induced eating through the integration of affective processes such as those specified in affect regulation theory (Heatherton & Baumeister, 1991).
Many participants had also indicated an awareness of the purpose and automaticity of their eating, though did not consider such awareness as sufficient to avoid engaging in unhealthy, overeating behaviour.Such a finding highlights the potential value of targeting stress-induced through strategies like implementation intentions where individuals can formulate specific plans to change their behaviour, accounting for automatic and reasoned processes behind behaviour (Keech & Hamilton, 2022;Rackemann, Hamilton, & Keech, 2024).In support of this, there was a small, significant effect of participants' action planning increasing following formation of their implementation intentions; further, majority of participants did not identify any difficulties in completing the implementation intentions exercise, with some also describing that the exercise increased their awareness and accountability for their behaviour.This suggests creating plans linking personal cues with an individual's preferred coping strategy may be useful to increase intention and motivation for behaviour.

Acceptability of replacement coping strategies
This study provided valuable insight into the various factors contributing to the perceived utility of replacement coping strategies for stress-induced eating.Participants described a range of cognitive and affective attitudes influencing their perceptions of a given coping strategy's utility.The most common cognitive attitudes reflected general beliefs regarding a strategy's utility, whether a strategy would only work temporarily, and further, whether a strategy would be utilised as a distraction or avoidance technique.Additionally, common affective attitudes pertained to whether participants considered that a strategy would increase or decrease negative or positive affect.Such behavioural beliefs may be vital to target within interventions seeking to promote healthy coping behaviours, given that previous research has supported the predictive utility of attitudes for behavioural intention across a range of health behaviours including food choice, and more recently, problemfocused coping (Keech & Hamilton, 2022;Riebl et al., 2015).
Participants also reflected upon the potential facilitators and barriers (control beliefs) to their engagement in each coping strategy.A key facilitator considered by participants was successful past and/or current experiences with utilising the strategy to reduce stress and/or stressinduced eating.This finding is not surprising, given that past behaviour has been indicated a key determinant of future behaviour (Keech & Hamilton, 2022).Dual-process theories suggest that behaviour is in part determined by learned associations and schemas formed from past experiences with stimuli (Rackemann, Hamilton, & Keech, 2024;Strack & Deutsch, 2014).An individual's perception of a strategy's utility may then be influenced by the existence of positive and negative beliefs stemming from their past experiences with the strategy, which in turn may influence the likelihood that a strategy will be perceived as effective in replacing an individual's habitual eating response to stress.
Barriers to engaging with a strategy were also considered by participants, including accessibility of the strategy-in terms of the time participants had available and the environment they were in-and levels of motivation and/or energy to engage with a strategy.Further, characteristics of the stressor itself were considered-such as intensity, type (i.e., changeable or unchangeable) and duration-which were described by participants as likely to determine whether the strategy would be useful or not in a given situation.Given the extensive range of stressors identified by participants as prompting their stress-induced eating, it follows that there would be variations in strategy acceptability amongst participants due to their most salient recent stressors.For instance, a university student considering an upcoming exam may consider situation modification as a beneficial coping strategy where an individual grieving a loved one may not; this highlights that different stressors may be best managed by either problem-or emotion-focused coping, or the application at both across time (Carver & Scheier, 1994).It is notable that normative beliefs played a less salient role and were only present in participant reflections of 'seeking support' as a strategy for replacing stress-induced eating.This may be because the ability of this strategy to reduce stress is reliant on another individual, and this is likely to be less so for the other strategies included within this study.
It was also common that participants indicated awareness of the automaticity and habitual nature of their eating behaviour, acknowledging that there would be a need for sufficient planning ahead and conscious decision-making to decrease engagement in stress-induced eating.This finding was largely congruent with previous research suggesting that coping behaviour tends to be habitual, highlighting the value in considering the automatic processes behind coping behaviour (Connor-Smith & Flachsbart, 2007;Keech & Hamilton, 2022;Lazarus & Folkman, 1984).Further, dual-process theories suggest that automatic processes take precedence over reasoned actioning of behaviour in times when cognitive capacity is under high load, such as in times of stress (Evans, 2008;Strack & Deutsch, 2014).As such, regardless of possessing knowledge and awareness of more adaptive coping responses to stress, individuals may not always make reasoned decisions to implement a given strategy when stressed if they lack motivation and ability (Keech & Hamilton, 2022).This suggests there are likely to be benefits in allowing individuals to choose their replacement coping strategies, where they can select the strategy perceived as most personally relevant and useful to their own situation.

Strengths, limitations, and future directions
The current study contributed to existing literature on stress-induced eating by providing a rich understanding of the factors individuals describe as important for them when determining the acceptability of a range of replacement coping strategies.This information can be utilised in the development of future interventions.Specifically, information about replacement coping strategies may be framed to target attitudes, barriers, and facilitators.Further, this study identified an extensive range of cues contributing to participants' engagement in stress-induced eating, providing an understanding of the motivating factors behind this behaviour that could assist interventions in aiding individuals to formulate specific plans for changing their coping behaviour.
A notable limitation was the representativeness of the sample; the sample consisted of community members, including university students (of which 61.48% were recruited via a subject pool).This may have resulted in a predominantly well-educated sample.Further, despite the sample population having a broad distribution in terms of weight and age, it was limited in that the sample was 88.37% female.However, previous research has suggested men do not typically respond to stress with indulgent eating behaviours, which may indicate that stress responses in men reflect different eating behaviours than women (Kim & Jang, 2017).Such a finding may prompt future research to investigate male populations and the potential changes in their eating behaviour.Additionally, those who self-reported reported never engaging in stress-induced eating in the past week were excluded from participation.While this procedure was followed to ensure the intervention would have current relevance for participants, we cannot rule out that excluded participants never engage in the behaviour at all, and therefore, the potential impact of this on the heterogeneity and representativeness of the sample should be considered in interpreting the results.
It should also be noted that the study is a feasibility and acceptability study, and conclusions cannot be drawn about the efficacy of the online tool.It is recommended that the tool be refined based on the feedback ascertained in the current study, and then be tested for efficacy in a randomised controlled trial.Consideration should be given to the range of potential replacement coping strategies included to address individual preferences.It is notable that in considering a replacement coping strategy's utility, participants in the current study indicated that any given strategy may only be useful in some situations and not others, and further, they reflected upon the impact of their own personal preferences and situations.Participants also commonly reflected upon how they would implement a replacement coping strategy, and this differed between individuals-for example, physical activity for one individual may involve running, whilst another would consider brief walks.This suggests it may be beneficial for future studies testing this online tool, to state that the suggested replacement coping strategies will vary in execution between individuals, and could consider prompting participants to write down detailed plans of how they will implement a replacement coping strategy when stressed.Last, some participant feedback indicated uncertainty around the likelihood of carrying out their plan when stressed.Future research may then also involve exploring intervention strategies able to bolster individuals' self-efficacy-such as self-monitoring-around implementing new coping strategies learned within interventions, to encourage lasting behaviour change.
It is also notable that we observed a small to medium effect size for changes in action planning from pre-to post-implementation intentions exercise in the current study.When comparing the effect size from the current study to the large effect observed by Rackemann, Hamilton, & Keech, 2024, it appears that the overall extent to which participants report having a plan to use their intended coping strategy is lower when there is greater heterogeneity of the behaviours being planned for.Therefore, it is important that future research in this area seeks to determine the differential effects of implementation intentions when various coping behaviours are being planned for, and magnitude of effect required to constitute a practically meaningful effect.

Conclusions
This study identified a range of specific internal and external cues for stress-induced eating, which can inform the content of future interventions.The acceptability of adopting an alternative coping strategy was described by participants as being determined by social cognition factors, including specific cognitive and affective attitudes, normative beliefs, control beliefs, automatic engagement in coping behaviour, and planning.Future coping interventions can consider these factors when providing information about new coping behaviours.The online tool significantly increased participants action planning for their chosen alternative coping strategy, and participants' descriptions supported that the use of implementation intentions for adopting alternative coping strategies to stress-induced eating is feasible and acceptable.However, plan quality was variable, and future interventions should provide further instruction on specifying a specific situation, strategy, start time, and duration.Representativeness of the sample (high proportion female, educated, students) was a limitation.Future research should refine the online tool and conduct a preregistered randomised controlled trial in a more representative sample.

Table 1
Online tool components and supporting research underpinning tool design.

Table 2
Participant cues identified for engagement in stress-induced eating.

Table 3
Themes of behavioural beliefs across replacement coping strategies.

Table 4
Themes of control beliefs across replacement coping strategies.

Table 5
Themes of automatic and volitional processes across replacement coping strategies.