Transforming nutrition by harnessing behaviour change theory for digital success
Adherence to advice and recommendations provided by digital tools is notoriously poor. One possible reasons for this is the poor integration of behaviour change theory. In this post we cover the state of the art on behaviour change and how this relates to the success of digital nutrition solutions. We also propose the use of behaviour design thinking as a novel approach for new product development and innovation.
Introduction to Behaviour change
Behaviour change is the conscious process of making decisions and acting on these. Whilst still considered a relatively new and niche area, behaviour change is a scientific discipline which has gathered steam over the last decade owing to growing evidence on the importance of understanding individual behaviours that can be explained by behavioural theories in order to target or integrated interventions that can impact health.
Relevance of behaviour change in health
It has been established that our behaviour contributes over 45% towards our health (WHO, 2009). That is more than genetics which contributes only around 20% (DHHS). This means that chronic illnesses such as Diabetes and Obesity are largely a result of our behaviours and mostly preventable. Most of the work that leverages behavioural science has been in the public health domain. These have focused on smoking cessation, improving public health, increasing physical activity and healthy pregnancies (Michie et al 2017). As technology has advanced, and consumer perception towards health has shifted towards prevention, so have public health initiatives, incentives and campaigns to nudge individuals towards adopting healthier habits. Understanding how individuals make decisions and how to support these decisions with relevant products and solutions is therefore key in shifting healthcare from reactive to proactive.
Behaviour change in Digital health
Rapid advances in science, technology, computing power and an increased focus on precision medicine have led to the growing area of digital health. The goal and promise of digital health is to:
- Increase access to information and care
- Lower cost of health care
- Increase self-management
- Provide 24/7 support
- Leverage new technology to provide personalized care
Examples include Artificial intelligent chatbots that can triage call to emergency rooms; blood pressure monitors that can measure blood pressure using a chest band and sending data to the general practitioner, digital therapeutics that can be prescribed by a Dr without the need for medication for specific conditions such as migraines (Perfood) or Irritable Bowel Syndrome (Cara Care).
The effectiveness of digital tools on Nutrition and behavioural outcomes
The nutrition and wellness industry have not been left behind in the technological era. In fact, a recent survey found that there are around 400K nutrition and health apps combined in Apple and Google play stores. Digital Nutrition solutions can range from wellbeing such as tracking food intake, all the way to medical nutrition such as those supporting the management of Diabetes. Behaviour change is inextricably linked to nutrition as previous research has demonstrated the impact that individual behaviours and psychology can determine outcomes of interventions.
Weight loss
For example, individuals who use digital tools such as a food diary lost significantly more weight (Patel et al., 2021), however the use of apps for weight loss in the long term is less effective (Chew et al ., 2022).
Similar findings were reported by another group who concluded that apps-based mobile interventions have a positive impact on the management of chronic diseases especially weight loss (El Khoury et al 2019)
Eating & physical activity
In an opinion paper published by European nutrition experts, concluded that depressive symptoms, emotional distress, fatigue and anxiety affect nutritional and behavioural outcomes such as eating and physical activity (Kohlenberg- Muller et al ., 2019).
Diabetes
In another service review study conducted by Diabetes and weight management platform Oviva, researchers found no difference in the weight loss achieved and acceptability of the service between face-to-face vs digital support (Huntriss et al., 2021) which indicates that digital offers a convenient and equally effective alternative to in-person care.
Body Mass Index (BMI) & Fruit & vegetable intake
Based on a recent systematic review and meta-analysis of 41 studies by Villinger et al 2019, researchers found beneficial effects of app-based mobile interventions on health-related outcomes such as BMI (Body mass index and blood pressure), clinical parameters (such as blood lipids) as well as a significant effect on fruit & vegetable intake. Whilst the effect sizes of the app-based mobile interventions were generally small (0.21 - 0.3), this points to the potential of integrating app-based mobile interventions as part of the nutrition care process.
Food as medicine
To date, most clinical effectiveness has been demonstrated in the Food as medicine approach, where individuals living with chronic conditions having received specific interventions to modify their behaviour. (Doyle et al., 2024)
Why does it matter? - The potential of leverage behavioural science to improve existing or develop new Personalized nutrition solutions for brands.
The reality is that while global smartphone penetration is high, the adoption of nutrition apps with the goal of improving health remains low at around 7 – 10% (Statista). Furthermore, for those who have adopted apps to self-monitor or improve their health, adherence is generally poor.
Behavioural science gets to the crux of why and how individuals make decisions or behave in a certain way. It is important for brands to understand the triggers, motivations and desires of why they are looking for a particular solution and the context in which those specific choices are being made as outlined above. Understanding the end-user at a deeper level can drive the development or incorporation of specific features that result in the desired behaviour and health outcomes.
For example, recent studies have found that ambivalence towards personalized nutrition can lead to a reduced uptake of the service, meaning that if consumers are not clear that an investment would lead to definite value or benefit, they are unlikely to demonstrate interest (Reinders et al 2020). In another study, researchers found that contextual factors influenced impact of personalised nutrition advice (Bouwman et al., 2022) whilst other researchers found that psychological characteristics determined their preference for the frequency and intensity of receiving personalised advice (Dijksterhuis et al., 2021).
A recent systematic review demonstrated that barriers to and facilitators to using nutrition apps can be grouped into 4 main reasons.
- The Individual – such as lack of motivation or capability
- The Technology – eg incomplete food databases privacy concerns or poor UX/UI
- The intended outcomes – such as goals, behaviour and emotional outcomes
- The social environment – such as social support and recommendations
Solutions that therefore best match an individual's preferences, (health) goals, expectations and capabilities can lead to higher satisfaction; higher engagement and better adherence to the advice provided. This will undoubtedly lead to higher revenue and better data generation, as well as a bigger social impact.
Yet, despite the potential of incorporating behavioural science into digital solutions, many companies do not know how to get started, or do not have access to domain experts to execute. This results in behaviour change being integrated as an afterthought rather than a core focus during ideation. Including behavioural assessment grounded in theory at the start, can lead to increased adherence and improved eating behaviour (Gibbs H & Chapman-Novakofski 2013:2012). This gap in design and deliver can be addressed by adopting Behavioural design thinking principles.
In addition, the industry has moved on from just product-market fit to ones that achieve product- behaviour fit.
In the next section we will provide a primer on Behaviour change theory, techniques and its role in digital health innovation.
Behaviour change theories
Behaviour change theory provides the “why” of human behaviour. As human behaviour can be very complex and contextual, a number of behaviour change theories have been developed that can be used as a basis to develop, explain or predict behaviour. These behaviour change theories can provide an actionable framework for developers and domain experts can use to develop new solutions and support individuals at every phase of their health journey in a way that it is effective and measurable (Kohlenberg-Muller et al., 2019).
In practice, behaviour change theories are a useful tool to explain behaviour at the beginning of a new innovation process in order to best match features and solutions to support individuals towards the intended behaviour change.
A summary of the commonly used theories is outlined below.
Behaviour change techniques in Personalized nutrition
Behaviour change techniques (BCT’s) are specific tools used to make the advice or guidance to make the desired health goals actionable. To date, the international standard nomenclature (terminology) used was developed by researchers in the United Kingdom Prof Susan Michie et al (2016). This set of 93 individual BCT’s have been grouped in specific Behaviour clusters or strategies.
The BCT’s can be mixed and matched in order to provide a better solution and also to improve user engagement and satisfaction. However in practice, the integration of BCT in commonly available apps is disappointingly very limited.
Based on a recent study including 41 studies that incorporated BCT’s into their app-based mobile interventions, the most common BCT’s used nutrition solutions tended to be goal & planning, feedback & monitoring, social support and shaping knowledge (Villinger et al., 2019)
Based on our most recent data of over 600 companies in the marketplace and on the Qina platform (April 2024), we found a similar pattern only that “shaping knowledge” came in 8th.
This demonstrates that overall, there is low integration of BCT’s in current solutions which could contribute to the poor adherence rates.
Another interesting evolution in the market is the use of biological data as a behaviour change tool with feedback as the BCT. This can include nutritional metabolites, continuous glucose measurements (CGM) or even, ketone measurements. Based on a recent systematic review of 767 articles, most RCT’s were aimed at Diabetes (30%), Cardiovascular (22.8%) and obesity (15%) with the most common biomarkers including anthropometry, blood pressure and glucose. The most commonly targeted behaviour was diet, physical activity and smoking.
The use of biological data for biofeedback is still early and in the absence of an existing framework on how to implement biological feedback research will be lagging (Richardson et al 2023).
To illustrate how BCT’s and Behaviour change theory work together, we provide examples in the scientific literature next.
Behaviour change theory and techniques in Personalized nutrition solutions
A number of research projects have used behaviour change as their theoretical framework to design interventions. We illustrate a short list of recent studies in the table below.
How do digital features relate to behaviour change effectiveness?
In order to make behaviour change techniques actionable, these need to be linked to features within the app or platform. For example, as mentioned, goal setting as a commonly employed behaviour change strategy may be incorporated in the form of a survey or diary feature. For feedback and monitoring, these may include food diaries, image recognition, chatbots or automated feedback via reports. There is currently no shortage of features available, but selecting the ideal BCT’s and combining different ones is where there is a dearth of research and information in terms of effectiveness. According to the literature, the most common features in effective solutions include:
- Food and activity tracking
- Social media
- Social support
- Feedback and monitoring
Based on the Qina platform the top 10 features included in apps are:
A recent systematic review of 19 studies (van Rhoon et al 2020) attempted to identify the interventions that were effective in producing clinically significant weight loss by linking the BCT’s and digital features. Researchers found:
- Social cognitive theory was the most common used BCT (n = 14),
- 7 BCT were identified in at least 75% of effective interventions.
- Researchers concluded that the inclusion of a larger number of behaviour change techniques were more effective in resulting in significant weight loss.
- In addition, social support and adding to the environment (eg trackers) were the most effective BCT’s in the short term whereas problem solving was most effective in the long term (van Loon et al., 2023).
- 3 digital features: shaping knowledge; diet tracking and activity tracking were frequently found to be effective.
- Social support was used in 100% of solutions but social media only in 50% of effective interventions suggesting that online support and face to face were equally effective.
In essence, developers who are looking to develop solutions need to have an understanding of which behavioural theory, behaviour change techniques to incorporate, and how to match these against Digital features that are most effective in the short and long-term based on the available literature. This is particularly relevant when using AI training data sets as well as continuously reviewing recommendations provided.
Market overview
The topic of behaviour change is particularly relevant as it relates to reimbursement, product design, health outcomes and research.
For example, ensuring significant weight loss in individuals living with Diabetes drives reimbursement. However, a recent study demonstrated the failure of digital programs in Diabetes to deliver significant outcomes (Peterson Health Technoloy, 2024)
Areas such as metabolic and women’s health are growing rapid owing to increased consumer interest and increased access of continuous glucose monitors (CGM’s) where real-time blood glucose is used as biofeedback. Wearable devices, sensors and monitors generate user-data that can be leveraged to provide personalized recommendations and advice for behaviour change (Erhard 2020). However, current solutions lack a theoretical approach, risking delivering well added-value and long term adherence.
Food as medicine programs have managed to combine expert-led education programs, virtual coaching with incentives and reimbursement to at-risk groups. At present, the question remains around the ROI of these programs, plus the standardized nomenclature used in the current body of evidence is limited.
Obesity management is very complex and requires long-term behaviour change, however preferences, traits, literacy levels, motivation and biological differences can vary substantially, meaning that identifying behavioural persona’s would be more relevant to ensure adherence.
Functional wellness is another market trend driven by Millennials and Gen-Z who are looking to consume foods and beverages with ingredients that contribute towards their health. For companies, it is important to understand how this groups perceives, understands and acts on their decisions around their health and how brands need to tap into this mindset (Innova)
Current early market players have leveraged behavioural scientists and techniques to deliver better value and experience. Furthermore, they have published the efficacy of their approach in terms of nutritional, health and behavioural outcomes. A few examples are provided next.
Current market leaders and innovators
Zoe is a UK based holistic Personalised health tech company that has developed an app. The solution starts a DIY kit which includes a Blood test, gut test, CGM, challenge test (cookies) and a questionnaire. A metabolic score is calculated based on the results which provides a guide to suitable foods and ingredients to consume.
BCT used: Biofeedback, goals & planning, shaping knowledge, feedback & monitoring, social support, incentives & rewards
Outcome: Zoe methods study (unpublished) 18-week study leads to improved biomarkers (ApoB, Cholesterol), nutritional parameters (waist circumference, weight) (Bermingham 2023)
January AI - A US based health tech company targeted at individuals living with metabolic dysfunction, PreDiabetes, Diabetes and healthy individuals. The solution involves wearing a CGM for 10 days, logging food and physical activity using the app. A metabolic score is calculated to guide individuals to the most suitable foods and ingredients to improve their metabolic health. A chatbot provides personalised recommendations based on data logged by the user.
BCT: shaping knowledge, goals & planning, biofeedback & monitoring, social support, coaching
Outcomes: A 20% improvement in Time in range (TIR). Based on their recent study, following the 10 day program, total calorie consumption reduced by 23% in the healthy group and 21% in the Prediabetic group couped with a increase in fiber intake.
Noom: A US-based digital company that delivers weight loss program using Cognitive behavioural therapy. Users log their food, activity and mood and receive personalized feedback and support via chatbot, expert support and recipes.
BCT: coaching, feedback & monitoring, goal setting, social support,
Outcome: Using natural language processing, unstructured text was analyzed to unravel how language used by customers can predict weight loss.
Outcomes: Users with goal striving language (long-term) lost significantly more weight than users who used goal setting (immediate) language
Insidetracker - A holistic health company that leverages blood (and other) biomarkers to deliver personalized insights and advice to improve Healthspan. Users take a blood sample at a partner lab. Blood results are delivered via the Insidetracker platform and app. Personalized dietary advice and recommendations are provided via the platform. Users can track their progress and biomarkers as well as integrate their wearables such as Applewatch and Fitbit data.
BCT: Biofeedback, goal setting & planning, shaping knowledge, social support
Outcomes: Users who had biomarkers that were out of range at baseline, experienced a trend towards the normal range after adopting the advice and recommendations.
Other companies who are staring to leverage advances in behavioural science include Levels who have recently launched a new feature “Habit loops” which allows users to track their goals, biomarkers and behaviours.
What it means and what needs to happen
The lack of industry-wide low adoption behavioural approaches means that current players are missing a great opportunity to improve adherence and increase loyalty. Behaviour change remains a low priority for companies who are focused on customer acquisition, when in fact, it can be a game-changer. It is clear that it is the well-funded companies who have widened their approach to include behaviour change, but adoption requires prioritizing and an understanding of the value that behaviour change can add to existing solutions. Improving current solutions requires a structured and strategic approach in order to demonstrate efficacy and a better customer experience.
What is needed right now:
- Operational – Adoption of a Behavioural Design thinking approach to design or improve products and solutions
- Organizational – Reviewing current solutions with a behavioural lens. Better informed, trained and diverse teams
- Better tools - Standardization of BCT nomenclature and data (variables) to track, analyze and measure effectiveness of BCT’s and how these relate to digital features
- Access to domain experts – who can guide ideation, development, monitoring and research solutions
What would happen if the above does not realize?
Recent consumer research has demonstrated that consumers are increasingly more proactive about their health. This has resulted in blurring of the lines between food/pharma/technology/lifestyle where behaviour is at the crux of any action taken.
The current risk is to continue to launch solutions that are tech-driven rather consumer-focused ie as an aid to support consumers along their health journey or towards their goals. Consumers will continue abandoning apps and solutions that do not meet their behaviour and lifestyle.
Companies will continue sitting on valuable behavioural data that can be analyzed to segment users, to improve current solutions with features that can add value to the user, or to create new ones.
The consumer dream of receiving real-time advice that integrates all their (non-invasively) collected data, will remain a distant dream as the standardized contextual data of what, when and why consumers make decisions will perpetually be missing.
The role of Behavioural design thinking in NPD
Design thinking is a non-linear and iterative process consisting of 5 stages that aim to understand users’ problems and to solve these with desirable products or features. The 5 stages include: understand, define, ideate, prototype and test (Figure 1). The goal of design thinking is to develop new solutions that meet 3 goals which are desirability, feasibility, affordability. Design thinking has been used across industries to innovate and renovate solutions and has pretty much overtaken traditional innovation processes. Design thinking has more recently entered the field of nutrition and behaviour change as a way of creating solutions that are consumer-centric and solves real-world problems.
Ethics and Behavioural design thinking
As AI becomes employed in all facets of our lives, nutrition and health is no different. At Qina, we believe that the ethics of AI is important to consider in the ideation, development and monitoring phase of new solutions that aim to target behaviour to optimize, manage or treat. We have published our perspectives and 7-pillar framework for action on this important topic in our white paper “The Ethics of AI at the intersection of nutrition & behaviour change”. The full paper can be accessed here.
Needless to say, ethics has become an important component of our Behavioural design thinking approach to companies who are looking to innovate.
How Qina helps companies and brands
We can help companies at all stages to adopt a Behaviour design thinking approach.
Our Qina process of behavioural design thinking is outlined below
By using a combination of domain expertise, our global network and a suite of digital tools we provide guidance, support and expert services with as much or as little as hand-holding as companies need.
We work with R & D teams in health tech, ingredients, food and nutrition. For more information or to discuss an idea, book a call by signing up here.
In conclusion
We are entering a new phase of digital health which is around the effectiveness of digital tools. As personalized nutrition becomes more popular, behaviour change has developed into the critical component that can lead to the success of a solution. Despite digital tools showing promise in delivering healthcare remotely, at present the integration of BCT’s is low, whilst at the same time research on effectiveness is sparse.
Success of personalized nutrition solutions lie in the development of new tools that incorporate behaviour theories, techniques and approaches in a way that can be easily identified, analyzed and monitored. In addition, companies can unlock value and leverage their existing data by understanding how, why and when customers make decisions using a theoretical and consumer research framework.
In order to ensure both societal impact and health effects of personalized nutrition tools, companies have to approach nutrition, behaviour change and technology simultaneously.
References
- Self‐Monitoring via Digital Health in Weight Loss Interventions: A Systematic Review Among Adults with Overweight or Obesity - Patel - 2021 - Obesity - Wiley Online Library
- Patel, M.L., Wakayama, L.N. and Bennett, G.G. (2021), Self-Monitoring via Digital Health in Weight Loss Interventions: A Systematic Review Among Adults with Overweight or Obesity. Obesity, 29: 478-499. https://doi.org/10.1002/oby.23088
- Huntriss R, Haines M, Jones L, Mulligan D. A service evaluation exploring the effectiveness of a locally commissioned tier 3 weight management programme offering face-to-face, telephone and digital dietetic support. Clin Obes. 2021; 11:e12444. https://doi.org/10.1111/cob.12444
- König LM, Attig C, Franke T, Renner B. Barriers to and Facilitators for Using Nutrition Apps: Systematic Review and Conceptual Framework. JMIR Mhealth Uhealth. 2021 Apr 1;9(6):e20037. doi: 10.2196/20037. Epub ahead of print. PMID: 34254938; PMCID: PMC8409150.
- Cosette Fakih El Khoury, Mirey Karavetian, Ruud J.G. Halfens, Rik Crutzen, Lama Khoja, Jos M.G.A. Schols,The Effects of Dietary Mobile Apps on Nutritional Outcomes in Adults with Chronic Diseases: A Systematic Review and Meta-Analysis,Journal of the Academy of Nutrition and Dietetics,Volume 119, Issue 4, 2019, Pages 626-651,https://doi.org/10.1016/j.jand.2018.11.010.
- Kohlenberg-Müller K, Ramminger S, Kolm A, Barkmeijer A, Gast C, Adam M, Le Bruyn B, Heine-Bröring R, Rachman-Elbaum S, Werkman A, Vanherle K, Höld E, Wewerka-Kreimel D, Valentini L. Nutrition assessment in process-driven, personalized dietetic intervention - The potential importance of assessing behavioural components to improve behavioural change: Results of the EU-funded IMPECD project. Clin Nutr ESPEN. 2019 Aug;32:125-134. doi: 10.1016/j.clnesp.2019.03.017. Epub 2019 Apr 26. PMID: 31221277.
- Gibbs H, Chapman-Novakofski K. Establishing content validity for the nutrition literacy assessment instrument. Prev Chronic Dis 2013. https://doi.org/ 10.5888/pcd10.120267. [33]
- Gibbs H, Chapman-Novakofski K. Exploring nutrition literacy: attention to assessment and the skills clients need. Health 2012. https://doi.org/10.4236/ health.2012.43019
- Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68
- Rankin A, Kuznesof S, Frewer LJ, et al. Public perceptions of personalised nutrition through the lens of Social Cognitive Theory. Journal of Health Psychology. 2017;22(10):1233-1242. doi:1177/1359105315624750
- Horne, J.R., Gilliland, J.A., O’Connor, C.P., Seabrook, J.A. and Madill, J. (2020), Change in Weight, BMI, and Body Composition in a Population-Based Intervention Versus Genetic-Based Intervention: The NOW Trial. Obesity, 28: 1419-1427. https://doi.org/10.1002/oby.22880
- Ho AS, Behr H, Mitchell ES, Yang Q, Lee J, May CN, et al. (2022) Goal language is associated with attrition and weight loss on a digital program: Observational study. PLOS Digit Health 1(6): e0000050. https://doi.org/10.1371/journal.pdig.0000050
- Bouwman, E.P.; Reinders, M.J.; Galama, J.; Verain, M.C.D. The Impact of Both Individual and Contextual Factors on the Acceptance of Personalized Dietary Advice. Nutrients 2022, 14, 1866. https://doi.org/10.3390/ nu14091866
- Dijksterhuis GB, Bouwman EP,and Taufik D (2021) Personalized Nutrition Advice: Preferred Ways of Receiving Information Related to Psychological Characteristics.Front. Psychol. 12:575465.doi: 10.3389/fpsyg.2021.575465
- Michie S, Richardson M, Johnston M, et al. The Behavior Change Technique Taxonomy (v1) of 93 hierarchically clustered techniques: Building an international consensus for the reporting of behavior change interventions. Ann Behav Med 2013; 46: 81–95
- https://phti.com/assessment/digital-diabetes-management-tools/#summary-of-findings
- Richardson KM, Jospe MR, Saleh AA, Clarke TN, Bedoya AR, Behrens N, Marano K, Cigan L, Liao Y, Scott ER, Guo JS, Aguinaga A, Schembre SM
Use of Biological Feedback as a Health Behavior Change Technique in Adults: Scoping Review J Med Internet Res 2023;25:e44359 doi: 2196/44359 - Doyle J, Alsan M, Skelley N, Lu Y, Cawley J. Effect of an Intensive Food-as-Medicine Program on Health and Health Care Use: A Randomized Clinical Trial. JAMA Intern Med. 2024;184(2):154–163. doi:10.1001/jamainternmed.2023.6670
- WHO 2009
- Peterson Health Technoloy institute, 2024 https://phti.com/assessment/digital-diabetes-management-tools/#summary-of-findings
- Westerman K, Reaver A, Roy C, Ploch M, Sharoni E, Nogal B, Sinclair DA, Katz DL, Blumberg JB, Blander G. Longitudinal analysis of biomarker data from a personalized nutrition platform in healthy subjects. Sci Rep. 2018 Oct 2;8(1):14685. doi: 10.1038/s41598-018-33008-7. Erratum in: Sci Rep. 2018 Nov 8;8(1):16778. PMID: 30279436; PMCID: PMC6168584.
- Perfood https://www.perfood.com