Why We Need to Evolve the Evidence Base for Personalized Nutrition and Functional Medicine
I deeply appreciate this important topic brought to us by Dr. Michelle Barrow as she answers questions put to her by our Nutrition Programs Director, Romilly Hodges. This exchange was prompted by Dr. Barrow’s recent peer-reviewed publication – Transforming Personalised Nutrition Practice – a worthwhile read which came across Romilly’s desk earlier this year. After all, personalized nutrition is a core component of the Functional Medicine toolkit. As a profession (of both functional medicine and personalized nutrition practitioners), we need to bridge the gap between currently-accepted evidence models and what is actually needed to support our practice. We need to ask the hard questions; then take the steps to build a better model. It’s something I have proposed before, and I am absolutely delighted to continue to champion. – DrKF
Why We Need to Evolve the Evidence Base for Personalized Nutrition and Functional Medicine
Romilly Hodges: What is personalized nutrition and how does it differ from other nutrition approaches?
Dr. Michelle Barrow: The term ‘personalized nutrition’ is somewhat challenging from the get-go since it does not yet have a single universally-accepted definition or approach. It is sometimes used when referring to (typically standardized) interventions that are personalized to an individual’s needs, also known as person/patient-centered care. At other times to interventions targeted at disease categories, such as interventions targeted to groups of patients according to their genetic and/or genomic type, also known as stratified evidence-based nutrition or precision nutrition. However, these definitions are arguably not comprehensive enough.
Personalized nutrition should encompass all of these meanings, including person-centered care, genetic and genomic considerations, and also a broad range of other functional, environmental and lifestyle considerations. And, perhaps most importantly, a concept that has been referred to as ‘pathophysiological reasoning.’
Let’s dive into what that means. Pathophysiological thinking considers both normal physiological mechanisms as well as their underlying signs, symptoms, contributing factors and connections between different body systems. It looks at biochemical interactions, pathways and biomarkers. In clinical practice, a pathophysiological reasoning approach takes the patient’s signs and symptoms, as well as health history, family history, environment and lifestyle factors across a lifespan, and organizes them under physiological categories, such as the headings on the functional medicine matrix map. By organizing the patient’s signs, symptoms and health issues into these physiological categories, and observing how they cluster together, the practitioner is able to identify potential mechanisms and imbalances contributing to the person’s disease or health issues. Practitioners can then prescribe interventions (e.g. anti-inflammatory diet) whose mechanisms of action aim to improve the mechanisms of pathophysiology (e.g. inflammation).
This is in contrast to the differential diagnosis clinical reasoning method used in mainstream healthcare and dietetics, where the patient’s signs and symptoms are used to identify diagnosis of illness, dysfunction and disease. Once a diagnosis has been established, clinical guidelines are used to form evidence-based interventions.Person-centered or stratified interventions may be used. However, clinical decision making based on this approach is criticized for being disease-centered, and for missing the unique factors that may be generating a particular condition in a particular person.
The net effect of using a pathophysiological reasoning approach is that we can be much more effective at understanding the unique mechanisms that contribute to each individual’s health presentation. And by addressing those in a personalized way we can generate better health outcomes.
Romilly Hodges: Why do traditional scientific methods, such as randomized controlled trials and prospective cohort studies, fall short in evaluating and guiding personalized nutrition and lifestyle interventions?
Dr. Michelle Barrow: Traditionally, randomized controlled trials and prospective cohort studies are considered to be the highest grades of evidence that we have to determine the efficacy of medical and nutrition interventions (especially when combined as meta-analyses and systematic reviews). This research typically provides summary statistics on the efficacy and expected variation of an intervention in a specific population group which is of course helpful in learning more about specific interventions.
However, the problem with the data that we get from randomized controlled trials is that they only provide answers on the average efficacy of an intervention in a population group. They don’t tell us whether an intervention works for this one specific individual that has sought our care, and so they limit the potential to provide truly individualized, validated care. Nevertheless, these studies are used to inform public health policy and develop standardized clinical guidelines.
Simply applying standardized, clinical guidelines to personalized nutrition practice (without adding pathophysiological reasoning) is always going to be problematic. Individual disease expression is by its very nature a hugely complex mix of genetics, environment, lifestyle and diet. While personalized nutrition practitioners do use the important data they get from these kinds of studies, they also combine it with pathophysiological reasoning in order to make a more accurate determination of the efficacy of that intervention in the individual they are working with.
Romilly Hodges: What kinds of research models make more sense for supporting evidence-based personalized interventions?
Dr. Michelle Barrow: To build out a broader evidence-base for a personalized nutrition approach we need to think differently. The development of a new, case-by-case evidence base, which uses statistical machine learning to make probabilistic predictions, would support clinical decision making in the personalized nutrition model and thereby increase positive health outcomes for patients.
These case data should then be statistically analyzed to assess, over a period of time, the effect of changing variables (diet, supplements, sleep, stress, etc.) on signs, symptoms, biomarkers and health outcomes. This should be pooled into a case-by-case evidence base which could then utilize statistical machine learning to predict the efficacy of personalized nutrition interventions and support personalized clinical decision making.
Romilly Hodges: What is the current status of this kind of evidence base?
Dr. Michelle Barrow: We are just at the conceptualization stage of building this out. The first priority is to develop and use robust, standardized and validated assessment tools that can be broadly applied across a population of personalized nutrition practitioners to gather consistent data: health questionnaires for signs and symptoms, health history, family history, genetics, environment, lifestyle, social life, diet, behavior and other factors which have an impact on physiological processes across a lifespan. Anthropometric measures as well as functional, genetic, microbiome, metabolomic, diagnostic and other biomarkers are included.
Some of these exist already and should be used where possible, e.g. a 3 day food diary, sleep scale, stress scale. I am also involved in developing a number of additional standardized, online clinical tools and questionnaires through the Centre for Nutrition Education and Lifestyle Management (CNELM) in the United Kingdom. We will start trailing the health history tool in clinical practice from July 2021. We will then look to validate these tools against biomarkers from laboratory analysis. And as we build out a database of clinical case data, we’ll also need independent studies to validate the statistical prediction models that will use algorithms based on the understanding of pathophysiological mechanisms of actions and the known associated interventions.
To generate sufficient and consistent data for analysis, practitioners will need to use these new standardized data gathering tools in their clinical practice. Real-time physical activity, sleep and food intake measures using smartphone applications in combination with comprehensive profiling will also be required. Collaboration with laboratories will enable the integration of patient test results.
This is easier said than done, of course, and will have to be nuanced. For instance, over time, it may be possible to develop long and short versions for some of the questionnaires, allowing practitioners to discuss, negotiate and agree with their clients which assessment methods to use at any given time. Practitioners will likely need an opportunity to build rapport before requesting completion of more invasive questions or tools. This approach should also be client-led; clients determine which range of non-compulsory tools or ongoing analysis they engage with at different stages of the consultation process. Clearly, only fully completed tools should be used for data mining and substantiating evidence to inform clinical practice. Evaluation of their impact on the practice process will also be required, and the analysis of multiple tools and questionnaires by functional medicine practitioners could be insightful.
Romilly Hodges: What is your ultimate vision for the future of personalized nutrition and its evidence base?
Dr. Michelle Barrow: I think there is tremendous opportunity here and that we have to make steps to address the gaps in personalized nutrition research collectively as a profession. The vision is that a case-by-case evidence base will enable not just structured case study data but longitudinal evidence from the outcomes of individual cases and groups of cases. Even beyond that there are many other avenues to explore:
An online data collection and data provision system, accessible for a range of health practitioners and consumers, should save practitioner time in researching and evaluating evidence-based recommendations.
Data collected from practitioners on their management approaches, for example, time spent on lifestyle coaching, teaching cooking skills, developing self-awareness, mindfulness, stress management skills etc. may allow further assessment of the influence of practice approach.
Time saved in evaluating evidence-based recommendations may be then better spent in incorporating coaching strategies alongside nutritional interventions to support behavioral change, which can positively influence client’s engagement, confidence and compliance with personalized programs.
Data from the database could not only inform patients and practitioners, it could potentially inform clinical pathology laboratories about the development of new tests and physiological functional assessment. Electronic tools could link to the evidence-based literature in numerous ways, including interpretive guides on test results provided by laboratories.
The bottom line: Without question, evolving the evidence-base for personalized nutrition practice to reflect interconnected pathophysiological reasoning is essential. The first step requires the use of robust data gathering tools and questionnaires. Even as the task may seem daunting, we need to start building out that infrastructure now. In time, these tools can integrate clinical and laboratory data and ultimately enable the development of a case-by-case evidence base that would allow for all-important data mining and knowledge discovery that supports not only the profession of personalized nutrition and the services that support it, but delivers the best health outcomes.
To find out more about the standardized assessment tools in development by CNELM, you can contact them at firstname.lastname@example.org.
About Dr. Michelle Barrow BSc, MSc, QTLS DProf.
Dr. Michelle Barrow, BSc (Hons), MSc, QTLS, DProf, is the Academic Team Director and Clinical Director at CNELM. Michelle is inspired to support them achieve their vision of leading the integration of personalised nutrition into healthcare services worldwide. She strives to develop the evidence base to support personalised nutrition practice through her academic work, research supervision, post-doctoral research, and publication. CNELM teaches online nutrition degrees including a BSc (Hons) Nutritional Science, MSc in Personalised Nutrition, and clinical courses such as the Personalised Nutrition Practice Diploma, through distance education. Michelle thoroughly enjoys supporting students to achieve their aspirations and goals. Her passion comes from overcoming her own health issues using nutrition. Michelle completed a Doctorate in Professional Studies (DProf) in 2019, titled “Leading transformation in Personalised Nutrition Practice”. Her doctoral research included the construction of clinical tools to enable the development of a new evidence base for personalised nutrition practice in obesity management. Michelle continues to work on the development of robust translational bioinformatics tools, using pathophysiological reasoning and systems biology approaches, as they are key to achieving evidence based personalised nutrition practice.
Barrow M, Bell L, Bell C. (2020) Transforming personalized nutrition practice. Nutr Rev. 2020;nuaa012. doi:
This content was originally published here.