Development and feasibility study of an app (Ladle) for weight loss and behaviour change

Introduction

Overweight and obesity have almost tripled worldwide since 1975, and in 2016 39% of all adults were overweight, and 13% were obese (Ng, Fleming & Robinson, 2014; NCD Risk Factor Collaboration, NCD-RisC; WHO, 2015). The highest rates of obesity are found in Tunisia, the USA, Saudi Arabia and Canada, and the lowest are found in China, Mali, Japan, Sweden and Brazil; the UK, Australia and New Zealand are all placed in the middle of the range (WHO, 2015). Overweight and obesity both have psychological and physical health consequences. They are associated with body dissatisfaction, low self esteem, anxiety, low mood and a general lack of confidence (Foresight, 2007; Pereira-Miranda et al., 2017). They also increase the risk of cardiovascular disease, heart attacks, diabetes, joint trauma, back pain, many types of cancer, hypertension and strokes, the likelihood of which increases as a person’s BMI increases greater than 25 (Foresight, 2007; Pereira-Miranda et al., 2017; Mokdad et al., 2004; Romero-Corral et al., 2006; Ortega, Lavie & Sui, 2017). In the UK the prevalence of diabetes increased from 46% to 56% between 1996 and 2005, which can be largely explained by the rise in overweight and obesity (Public Health England, 2015). Obesity is also directly linked with mortality and decreased life expectancy (Mokdad et al., 2004; Romero-Corral et al., 2006; Ortega, Lavie & Sui, 2017; Global BMI Mortality Collaboration, 2016).

Excess body weight also has economic consequences. The global economic cost of obesity is approximately $2.0 trillion, or 2.8 percent of global GDP, paralleling the global impact of smoking or armed violence, war, and terrorism (McKinsey, 2014). Further, the toll of obesity on health-care systems is between 2 and 7 percent of all health-care spending in developed economies which increases to 20% if the cost of treating associated diseases is included (McKinsey, 2014). In the UK, it is estimated that the NHS spent £6.1 billion on overweight and obesity-related ill-health in 2014 to 2015 which is higher than the combined cost of the police, fire service and the judiciary for the same period (Public Health England, 2017).

Obesity and overweight are therefore a health and financial burden. Many people are motivated to lose weight and studies indicate that about 42% of the general population report trying to lose weight in the past year; that about 23% report trying to maintain weight loss in the past year; and about 70% report having ever dieted to lose weight (Santos et al., 2017). Research has therefore evaluated the effectiveness of different weight loss interventions with a focus on face to face and digital approaches.

Face to face weight management programmes can be group interventions or involve one to one support with a health care professional. Both tend to draw upon a wide range of behavioural strategies such as self monitoring, reinforcement, cognitive restructuring, relapse prevention and nutritional information and involve health care expert input. In terms of effectiveness, a review of the evidence by NICE indicated that by one year, those who had received best case evidence based behavioural management from either public sector or private sector weight management behaviour change services such as the NHS, Weight Watchers or Slimming World showed an average weight loss by one year of 2.22 kg (Hartmann-Boyce et al., 2014; Johns et al., 2014). Furthermore, a large scale analysis of weight loss by Slimming World members showed an overall mean change of 3.9 kg and a % weight loss of −4.4% by 12 weeks with those who completed at least 75% of weekly sessions showing a mean weight loss of 6.8 kg and a % weight loss of −7.5% by 12 weeks (Stubbs et al., 2015). Likewise, a trial of either 12 weeks or 52 weeks of Weight Watchers sessions compared to a brief intervention showed mean weight losses by 12 months of −4.75 kg (12 wks group) and −6.76 kg (52 wks group) both of which were greater than the brief intervention (Ahern et al., 2017). Further a trial evaluating Slimming World versus advice illustrated 40% attendance rates at Slimming World, with a mean weight loss of 2.43 kg by one year follow up (Aveyard et al., 2016). Evidence based face to face interventions are therefore moderately effective in producing weight loss, but are costly due to the need for trained professionals to deliver the programme. For example, in the UK, Weight Watchers on referral costs the NHS £45 a participant (Ahern et al., 2011) which, if offered to the 42 million overweight people in the UK would cost the NHS approximately £2 billion. Further an estimate of the incremental cost of Weight Watchers indicated a cost of £159 per kg lost for 52 weeks and £91 per kg lost for 12 weeks (Ahern et al., 2017).

In contrast, digital interventions can be delivered at a lower cost than face to face programmes which is reflected in a recent proliferation of online resources and apps. For example, a scoping review in 2015 identified 393 weight loss apps from 4 commercial app stores (Rivera et al., 2016). Of these, however, only 3 had been scientifically evaluated and only 1 involved health care expert involvement. In addition, the majority were limited in their evidence base. For example, the scoping review concluded that although self-monitoring was the most commonly used approach (35.3%), followed by physical activity support (27.5%) only a quarter used weight assessment (25.4%) or healthy eating support (23.2%) and even less used goal-setting (21.4%), motivational strategies (7.1%), social support (5.3%) or personalized feedback (7/393, 1.8%) (Rivera et al., 2016). In terms of effectiveness, a systematic review and meta -analysis in 2015 identified 12 research studies evaluating apps for weight loss and concluded that the apps were more effective than control groups resulting in a mean of −1.04 kg greater weight loss (Flores Mateo et al., 2015). These apps are therefore less effective than evidence based face to face programmes (Hartmann-Boyce et al., 2014; Johns et al., 2014; Stubbs et al., 2015; Ahern et al., 2017; Aveyard et al., 2016) yet cost less and can be made available to a wider population.

One compromise approach is to provide an app together with human coaching to deliver behavioural strategies. This has been established in the US with the Noom app (Kim, Ray & Veluscek, 2017; Toro-Ramos et al., 2017; Chin et al., 2016; Michaelides et al., 2016). Analysis of this approach indicates greater weight loss compared to apps without coaching (−5.2% by one year follow up) but the employment of human coaches increases the cost of the intervention. For example, consumer access to the Noom course costs $59 (∼£46) per month (Noom, 2018) which is more expensive than consumer access to a Weight Watchers face to face programme where costs start at £17.95 per month (Ahern et al., 2011). In a similar vein, Little et al. (2016) evaluated the impact of combining an internet based intervention with nurse support which was provided either face to face or remotely by email and telephone compared to a control group who received dietician advice and a dietician follow up session. The results showed that both internet groups reported greater weight loss than the control group (−2.64 kg by 12 months) and that whereas those who received internet plus face to face nurse support reported a mean weight loss of −4.14 kg by 12 months those who received internet plus remote nurse support reported a mean weight loss of −3.94.

In summary, the obesity epidemic highlights the need for effective interventions that can reach a wide population. Whilst face to face interventions show some effectiveness they are expensive and limited in their reach. In contrast, apps are less effective but cost less and have wider reach unless human coaching is added which drives the cost up. Some of the key comparison interventions are shown in Table 1 although definitions, timing of interventions and outcomes measures vary between studies so any comparison is tentative.

There is therefore a need for a low cost digital intervention that can be made available to many but is as effective as a face to face programme. The aim of the present study was therefore to develop and evaluate an evidence based app that provides many of the evidence based behaviour change strategies contained within the face to face approaches, involves health care experts in its development but that is also low cost and available for general use by a wider population. This paper describes the development of the new Ladle app and its preliminary evaluation.

Development of the app

The app development team involved a Professor in Health Psychology with 30 years experience working in eating behaviour and weight management (JO), a registered dietician (SG), two psychology researchers who helped with scripting the habits and lessons (CW and DM), a trained voice artist (RMJ) who delivered the habits and lessons for the audio recordings and the two app developers who run the app company, invested in the app and project managed the whole process (HM, AW).

The process of developing the app followed lean start-up methodology (Ries, 2011). In line with this, the development team worked with clinicians and prospective end users of the app throughout the development process to ensure the Ladle app was developed to meet their needs. The steps of the app development were as follows:

Step 1: Needs assessment with health care professionals: Interviews were held with academic researchers specialising in obesity (n = 5), Tier 2 and Tier 3 Doctors specialising in obesity (n = 3), Consultant Clinical Psychologists (n = 2), Dieticians (n = 4), Clinical and Programme Leads for Digital Psychological Therapy (both NHS provision and private companies) (n = 4), and Clinical Commissioners for Diabetes (n = 5) to explore whether a digital, evidenced based, behaviour change app for weight management would be valuable. All those interviewed gave positive feedback.

Step 2: Needs assessment with overweight and obese patients: A Google advert describing the proposed Ladle app was set up which drove traffic to an online landing page where a form could be filled out to learn more once the app is ready. Results were 2.6% un-optimised CTR (Click Through Rate) and 6.3% conversion to EarlyBird product. This outperforms the average CTR on all Google Paid Ads of ∼2%, and the average conversion rate for Health and Medical products of 3.27% (WordStream, 2018) indicating that a digital, evidenced based, behaviour change app for weight management would be desirable to end users.

Step 3: Evidence review: The structure and content of the app was developed in line with three bodies of research evidence (see 25 for a review): (i) research exploring the predictors of weight loss and weight loss maintenance (Hartmann-Boyce et al., 2014; Johns et al., 2014; Dombrowski et al., 2014; Stubbs et al., 2011; Teixeira et al., 2015; Thomas et al., 2014); (ii) research assessing differences between those who lose weight and maintain this weight loss and those who do not using the National Weight Control Registry together with other research (Thomas et al., 2014; Wing et al., 2008; Wing & Phelan, 2005; Ogden, 2000; Elfhag & Rössner, 2005; Butryn et al., 2007; Epiphaniou & Ogden, 2010a; Epiphaniou & Ogden, 2010b); (iii) qualitative research exploring the characteristics of successful weight loss maintainers (Ogden & Hills, 2008; Epiphaniou & Ogden, 2010a; Epiphaniou & Ogden, 2010b; Greaves et al., 2017). This research highlights the key role of eating breakfast, planning meals, eating meals rather than snacks, avoiding denial and being more active which are reflected in the six behavioural habits. These habits are also reflected in the NHS Choices structure for weight management (NHS, 2018) although recent research has questioned the beneficial role of breakfast (Sievert et al., 2019). Furthermore, this research illustrates the importance of psychological strategies such as goal setting and planning; self monitoring, managing emotional eating; a cost benefit analysis of eating behaviour; cognitive restructuring; relapse prevention; recognising the meaning of food; peer support; eating mindfully; having a positive body image; positive reframing; reinforcement; recognising a behavioural model of weight gain; developing a new identity; and self compassion which form the basis of the 36 psychological lessons. This evidence review resulted in a draft course curriculum for the app.

Step 4: Focus groups to assess usefulness of proposed course content: Focus groups (n = 2) were conducted with overweight and obese individuals (n = 19) and interviews were carried out with spokespeople from obesity charities and obesity interest groups (n = 4) to gain feedback on the draft course curriculum. All feedback was positive and minor amendments were made in line with suggestions.

Step 5: Producing the habit and psychological lesson content: The app content was written to reflect the evidence review and focus groups. Two psychology researchers were employed to write the scripts for the app and a registered Dietician who specialises in obesity management wrote the recipe cards and created simple rules to avoid the need for calorie counting. This process was overseen by JO, AW and HM.

Step 6: Building the app: Software developers were employed to build the Ladle digital weight management course including all habits and psychological lessons. This process was overseen by AW and HM.

Step 7: Ongoing user feedback: Users were asked for their ongoing feedback to inform continuous improvement of the Ladle app. This involved either face to face or virtual meetings as well as written feedback within the app itself (n = 159 comments; see results for further details). Changes were made to both the structure and content of the app to reflect this feedback.

Results

Discussion

Obesity and overweight have physical and psychological health consequences (Foresight, 2007; Pereira-Miranda et al., 2017; Mokdad et al., 2004; Romero-Corral et al., 2006; Ortega, Lavie & Sui, 2017). They also have economic implications for the health care system (McKinsey, 2014). Existing behaviour change programmes either use face to face interventions which are moderately effective, but expensive and have poor reach or digital interventions which have wider reach, cost less but have limited if any effectiveness which is only improved with the availability of a human coach (e.g., (Hartmann-Boyce et al., 2014; Johns et al., 2014; Stubbs et al., 2015; Ahern et al., 2017; Aveyard et al., 2016; Rivera et al., 2016; Chin et al., 2016; Little et al., 2016). The present study therefore aimed to develop and evaluate a digital app based weight loss programme which incorporated the evidence based behaviour change strategies used by face to face interventions but was lower cost and available to the wider population.

The resulting Ladle app was developed using seven steps including expert, health care practitioner and service user input and evidence review. It involved a focus on 6 behavioural habits supported by 36 audio psychological lessons and took 12 weeks to complete. The app was evaluated in terms of completion rates, weight loss, behavioural habits and participant feedback.

The results from the evaluation showed that 75 people were considered starters and of these 44% (n = 33) were considered completers. This is lower than some of the studies reporting data for the completion rates for the commercial groups such as Weight Watchers and Slimming World and NHS groups who offer face to face group sessions (e.g., Ahern et al., 2011; Laws R; Counterweight Project Team, 2004; Logue et al., 2014). It is also lower than available apps that offer a human coach (e.g., Pagoto et al., 2018; Allen et al., 2013; Michaelides et al., 2016). It is, however, comparable or higher than a few of these more intensive interventions (Stubbs et al., 2015; Aveyard et al., 2016). No completion rate data is available for apps without a human coach. This suggests that although a digital app with no coach can have wide reach and attract a high number of users fewer people may complete the course than for a more intensive face to face weight management programme.

In terms of actual weight loss, the results were much more positive. In particular, 52% showed at least 5% weight loss which has been associated with improved health status (Wing et al., 1987; Aucott, 2008) which was comparable or higher than that reported by some face to face interventions (e.g., Ahern et al., 2011; Ahern et al., 2017; Laws R; Counterweight Project Team, 2004; Logue et al., 2014; Aveyard et al., 2016). It is also higher than some digital apps with a coach (Michaelides et al., 2016; Allen et al., 2013). In addition, 72% showed greater than 3% weight loss which also brings benefits, not only in terms of health but also as an incentive to persist with weight loss efforts. Furthermore, in terms of absolute weight loss (mean −3.8 kg; median −3.5 kg) or percentage weight loss (mean 4.65%; median −5%) Ladle was comparable to available data for several face to face interventions (e.g., Aveyard et al., 2016; Logue et al., 2014; Ahern et al., 2011) and an improvement on apps with no human coach (e.g., Turner-McGrievy & Tate, 2011; Laing et al., 2014).

The Ladle app was also evaluated in terms of 6 behavioural habits, namely healthy breakfast, lunch, dinner, healthy snacks and drinks, eating at planned times and being active which have been consistently linked with weight loss (Ogden, 2018; NHS, 2018). The results showed that by the end of the 12 week course the large majority of participants had adopted 4 of the 6 habits and nearly half had adopted the remaining two habits which should not only bring about immediate health benefits but set the scene for longer term behaviour change and weight loss in the future.

Finally the app was evaluated in terms of participant feedback which was mostly positive. In particular, participants were extremely positive about the content of the lessons. Their neutral comments described statements of commitment and any challenges they faced with adhering to the programme and the negative comments mostly reflected technical issues which were addressed throughout the evaluation. The app was therefore found to be acceptable to this patient group.

There are some problems with the current study that need to be considered. First the study utilised a single arm prospective design without randomisation and without a control group. This was due to the desire to collect preliminary data regarding completion rates and weight loss as a first stage to testing the app. Second, weight loss data was collected from within the app and was therefore self report which may reflect issues of social desirability and error. Further, comparisons with existing studies are problematic due to the vast variation in designs used, interventions, duration of interventions, definitions of completions rates and weight loss and whether the data were collected as part of a larger randomised control trial or a service evaluation. These comparisons should therefore be seen as tentative and a preliminary analysis. This present study, however, provides a detailed insight into the process of developing an evidence based app and some preliminary insights into its usefulness and impact on weight loss and habit change.

Conclusion

In summary, this feasibility study aimed to develop and evaluate a weight loss app that was low cost with a wide reach but based upon the evidence and utilising the evidence based behaviour change strategies associated with sustained weight loss maintenance. The Ladle app offers a 12 week course designed to promote 6 key habits and weight loss facilitated by psychological lessons. The results of the analysis indicate that although completion rates were lower than many face to face interventions and apps with a human coach, weight loss was comparable to several of these more intensive and costly forms of weight management and better than that that achieved by other apps with no human coach. This new app therefore offers an evidence based digital approach to weight management which is as effective as other approaches but can be delivered at lower cost to a wider population.

Supplemental Information