The expectations and realities of nutrigenomic testing in australia: A qualitative study

Abstract Background Consumer genomic testing for nutrition and wellness, (nutritional genomics), is becoming increasingly popular. Concurrently, health‐care practitioners (HPs) working in private practice (including doctors interested in integrative medicine, private genetic counsellors, pharmacists, dieticians, naturopaths and nutritionists) are involved as test facilitators or interpreters. Objective To explore Australian consumers’ and HPs’ experiences with nutrigenomic testing. Method Semi‐structured in‐depth interviews were conducted using predominantly purposive sampling. The two data sets were analysed individually, then combined, using a constant comparative, thematic approach. Results Overall, 45 interviews were conducted with consumers (n = 18) and HPs (n = 27). Many of the consumer interviewees experienced chronic ill‐health. Nutrigenomic testing was perceived as empowering and a source of hope for answers. While most made changes to their diet/supplements post‐test, self‐reported health improvements were small. A positive relationship with their HP appeared to minimize disappointment. HPs’ adoption and views of nutrigenomic testing varied. Those enthusiastic about testing saw the possibilities it could offer. However, many felt nutrigenomic testing was not the only ‘tool’ to utilize when offering health care. Discussion This research highlights the important role HPs play in consumers’ experiences of nutrigenomics. The varied practice suggests relevant HPs require upskilling in this area to at least support their patients/clients, even if nutrigenomic testing is not part of their practice. Patient or public contribution Advisory group included patient/public group representatives who informed study design; focus group participants gave feedback on the survey from which consumer interviewees were sourced. This informed the HP data set design. Interviewees from HP data set assisted with snowball sampling.


| INTRODUC TI ON
In recent years, personal genomic testing or 'consumer genomics' has placed genomic information into the hands of consumers. These tests promise risk predictions in diverse areas including fitness, response to medications and traits such as premature balding. Also popular are tests for nutrition and wellness (nutrigenetic and nutrigenomic tests, hereafter referred to collectively as nutrigenomic tests), which offer diet and lifestyle recommendations that are personalized to genetic predisposition(s) to metabolic responses. 1 Often advertised and sold online, these tests are marketed as being empowering and health transformative. 2 Clinical validity (strength of the disease-gene relationship) and clinical utility (impact on health outcome) are essential when evaluating any genomic test. Current Australian guidelines recommend against the use of nutrigenomic testing unless the test, and related dietary interventions, are sufficiently evidence-based. 3 Strong evidence does exist for certain gene-nutrient relationships, as is the case for diets low in phenylalanine prescribed to people with phenylketonuria, a genetic disorder of amino acid metabolism. 4 However, for commercially available nutrigenomic tests, many of the selected gene-nutrient interactions, and the associated dietary advice, lack sufficient evidence to be considered clinically valid and useful. 5,6 Communicating the importance of clinical validity and utility becomes a challenge when genomic tests enter the consumer realm. An example of this is MTHFR gene testing. 1,7 The MTHFR (methylenetetrahydrofolate reductase) gene is responsible for the production of an enzyme involved in folate metabolism. Many online sources implicate two MTHFR single nucleotide polymorphisms (SNPs) in the risk of developing a myriad of health conditions and support the validity of testing and the utility of various supplements for prevention and treatment. 8 However, a key international genetics organization 9 and prominent consumer genomics company 23andMe 10  nutritionists, are offering to facilitate testing or provide support regarding results interpretation. 2 Other HPs in private practice known to provide these services include general practitioners (GPs) with an interest in CAM ('integrative' GPs) and those working in allied health, such as dieticians, pharmacists and private genetic counsellors (GCs). 2 While GCs may not necessarily instigate this type of testing, in Australia there has been a substantial rise in consumer requests regarding consumer genomic testing, with 11% of genetic services reporting queries in 2011 13 increasing to 66% in 2017. 14 Additionally, in Australia, pharmacogenomic and nutrigenomic tests are sold by pharmacists over the counter in some community pharmacies. 15 Published practice guidelines for GPs caution against the use of this testing, 16 however, there is limited understanding of the practice of those in the fields of CAM and allied health. 17 Further, little is known about consumer experiences of nutrigenomic tests, including those for MTHFR. The Genioz (Genomics: National Insights of Australians) study sought to explore consumer views and, where appropriate, experience with a variety of consumer genomic tests. [18][19][20] Survey responses and follow-up interviews revealed that some Australians are interested in and are pursuing nutrigenomic testing, often via a CAM practitioner. 19,20 These findings informed subsequent research to explore the landscape of HPs offering and/or interpreting consumer genomic tests. This additional work was part of the Workforce and Education Program of Australian Genomics Health Alliance (Australian Genomics), which focusses on the workforce implications of genomics in Australia. 21 This paper presents the combined findings of these interviews with both consumers and HPs to explore their experiences with nutrigenomics in Australia.

| ME THODS
This study took a qualitative approach, comprising interview data from two data sets (Genioz to explore the experiences of the consumers and Australian Genomics to explore HPs' experiences). Ethics approval was obtained for both studies (HREC 1 545 806.3, Genioz project; and HREC 1 646 785.9, Australian Genomics). The Genioz project was overseen by an advisory group including representatives of patient/public groups who advised researchers on study design and assisted with recruitment.

| Participant recruitment
Consumers were purposively sampled from Genioz online survey respondents ( Figure 1). 20 Survey respondents who indicated they would be prepared to be interviewed were sent an email. Given that survey respondents self-selected the type of test they had pursued, screening calls took place to confirm eligibility. It became apparent that many potential interviewees had pursued single gene testing for MTHFR only. Therefore, as recruitment and data collection continued, additional individuals who had undergone testing for this gene only were excluded to ensure all aspects of the data set were well captured. 22 Specific HPs working in private practice were identified as those who do or could offer nutrigenomic testing (informed by findings of the Genioz study). 20 CAM (naturopathy, nutrition, integrative medicine) and allied health (dietetics, pharmacy and private genetic counselling) practitioners were recruited via four methods: purposive sampling after an online search to identify key individuals; random sampling through advertisements placed in practitioner-specific professional society e-newsletters; convenience sampling through professional research networks; and snowballing sampling through interviewee networks. Invitation letters were emailed, to which interested practitioners responded directly to the researcher.

| Data analysis
Audio recordings were transcribed verbatim and de-identified, with pseudonyms assigned to consumers and ID numbers assigned to HPs. NVivo 11 and NVivo 12 (QSR International, Melbourne, VIC, Australia) were used for data management, for consumers and HPs respectively.
For each data set, transcripts were coded by ET (consumers) and CH (HPs) and analysed thematically, using an inductive, constant-comparative approach. 23 A sample of the transcripts was coded independently by at least one other researcher for each data set (CH and SM for consumers; AR and SM for HPs), from which a consensus was reached regarding the final themes.
Coding frameworks were developed for each data set with their own set of themes identified. Themes were then compared by ET, CH and SM, imported into a new NVivo 12 file and coded further, resulting in common themes presented in this paper and agreed upon by the other authors (additional consumer themes were published in a Genioz Honours research thesis, and additional HP themes are to be reported in a separate manuscript). Coding was also conducted regarding HPs' views of and engagement with nutrigenomic testing in their practice. This resulted in a detailed matrix of the landscape of practice for all HPs, which was then visually represented as four quadrants. Four of the researchers (CH, BT, ET and SM) reached consensus with the placement of HPs in each specific quadrant based on each HP's views of, and level of engagement with, nutrigenomic testing described in their interview.

| Participant characteristics
Eighteen consumers ( Figure 1) and 28 HPs were interviewed ( Figure 2). Participant characteristics are presented in Tables 1 and   2, respectively. In the consumer cohort, HPs first suggested nutrigenomic testing to eight consumers, while 10 consumers initiated their own testing and sought advice post-test. While the HPs interviewed in this study were cautious about nutrigenomic testing, their views of, and practice around, nutrigenomic testing varied, as shown in Figure 3. Their views ranged from being sceptical ('testing sceptic': left quadrants) to enthusiastic ('testing enthusiast': right quadrants).
There was also a mix of engagement with testing provision: whether they adopted testing in their practice, including ordering a test and/ or interpreting test results ('clinical adoption': top quadrants), or reported little or no engagement ('no adoption': bottom quadrants).

For example:
The DNA test, it actually zones in a little bit more … I send them off to the chemist to get them…it made me go, oh great, I've got science answering my question.

| Themes
Three key themes were developed from the two data sets. Illustrative quotes are included in-text, with additional quotes in Table 3. 3.2.1 | Theme 1 -Nutrigenomics: offering a source of hope?
Thirteen of the consumer participants had a long history of health problems and were yet to find any answers or solutions to provide a diagnosis or to guide future treatment and management of their condition, as shown in Table 3: Quote 1 and also described below: And I've seen so many different doctors and I've done the standard tests, you know, just the standard blood tests a million times, only to get the same sort of response from my doctors that I was all good, it was all fine.
(Lucy) Some consumers were unsure what to expect from nutrigenomic testing and described 'taking a leap of faith' that testing would provide not only answers but tangible solutions to their health concerns ( her raw data to third-party interpretation site and took results to GP. She has made minor changes to diet. Each HP was coded as enthusiastic, sceptical or in between according to their views on testing. The same process was conducted for HP engagement with testing.

TA B L E 3 Additional quotes
Whereas other HPs spoke of the limitations of testing ( (Natasha) Angela and Bridget reported that, following the results of their first nutrigenomic testing, they were advised by their HP to purchase additional tests, including for ancestry. This was, as Angela explained, to gain 'more information'. Several consumers reported that they were advised to download their raw genomic data from ancestry genomic testing and use it to obtain health information from online third-party interpretation programmes (  A few HPs, particularly, private GCs, felt testing did not fit within their practice and would advise clients to go elsewhere (Table 3: Quote 10). However, many had adopted nutrigenomics in their practice in some way: actively ordering testing or encouraging clients to order nutrigenomic testing online and bring the results back for interpretation ( As a result of their testing, some consumers reported making minor changes to their diets and others changed their supplement regimens. However, despite the perception that nutrigenomic testing gave them answers ( Nevertheless, all consumers remained positive about their decision to pursue nutrigenomic testing and valued the fact that they had found an HP who was not only willing to work with their nutrigenomic result but also validated their health concerns. Thus, it appears that nutrigenomic testing had personal utility even in the absence of actionable results.

| D ISCUSS I ON
Nutrigenomic testing has grown in popularity in Australia, with interest from consumers and HPs. This may in part be due to the positively framed marketing that has circulated online. 24 It appears that these online marketing strategies, portraying nutrigenomic testing as 'health transformative' 2 particularly resonate with people who are chronically unwell and those with an interest in CAM.
Most consumers in our study reported their health as being 'poor' or 'fair' and described themselves as being chronically unwell.
Unlike early adopters of consumer genomic testing, who were primarily motivated by curiosity and an interest in the science, 25 the chronically unwell consumers in this study saw nutrigenomic testing as a 'means to an end', having tried everything else and finding no answers to their health concerns. This is consistent with a previous characterization of CAM users. 26 While nutrigenomic testing was not as health-transformative as the consumer participants had hoped, they remained positive about their decision to pursue testing. The majority of consumers reported feeling empowered and validated after receiving their nutrigenomic testing results. This was particularly true for those who discovered they had a particular MTHFR SNP that they felt could explain their health concerns.
Early qualitative research exploring patient and doctor experiences of diagnosing MTHFR polymorphisms and providing treatments noted the vast array of self-reported, unvalidated information online. 27 In one study from the USA, patients reported the frustration of seeing many doctors throughout their lives and receiving no relief to their chronic symptoms and the majority described receiving their MTHFR test results as validating, empowering and life-changing. 27 Likewise, some of the chronically unwell consumers in our study perceived MTHFR SNPs to be the diagnosis for which they were searching. After receiving feedback from 'mainstream' HPs that nothing could be done for them, when participants However, the general population may not derive the same perceived benefits from nutrigenomic testing and may instead be at risk of disappointment.
The trusting relationship the consumers had with their HP who was involved with the nutrigenomic test (or test results) also mitigated disappointment with the test outcomes. Much of the positive experience of nutrigenomic testing therefore came from finding an HP who was willing to listen and did not immediately discount their chronic symptoms. This has been similarly described in the area of homeopathy. 30 Given this, it is possible that the positive response to nutrigenomic testing relates less to nutritional genomic science and more to psychosocial factors including receiving support, empathy and validation from their HP. Having an HP who supports nutrigenomic testing, and has even had testing themselves, can add a sense of legitimacy to the test which may influence clients accordingly. HPfacilitated testing is also one way of mitigating the possibility that consumers will misinterpret their test results. However, the range of disciplines and the varied nature of practice between HPs could prove problematic for consumers in choosing an HP. Additionally, concerns remain that HPs may be inadequately trained to facilitate informed decision making. 31,32 Given that some nutrigenomic testing companies train the HPs in-house, 19 this represents a potential conflict of interest and could translate to consumers receiving biased information and advice based on poor evidence.
A key tenet of genetic counselling, now a regulated profession in Australasia, 33 is facilitating clients' long-term adaptation to genomic information. 34 Genetic counsellors have traditionally been employed in hospitals within a public health system, counselling patients for tests ordered via an accredited laboratory. Scepticism regarding validity and utility of nutrigenomic testing, accompanied by GCs' perception that commercial consumer genomic tests may not belong in the public health genetic clinic, is a valid concern. However, more recently, GCs are working in private practice 33

| Limitations
The majority of eligible consumers responded to the invitation for recruitment, however, given the nature of this process, it is likely not all types of experiences of personal genomics were captured in this cohort. Most consumer participants were relieved to discover an MTHFR polymorphism, but others who had a more negative experience may not have responded to the survey. Access to nutrigenomic testing may also vary based on ethnicity and socio-economic status.
Consumers in our study identified as primarily Australian (presumably Caucasian -noting that they could select more than one ethnic ancestry in the survey), highly educated, and resided in areas of middle to high socio-economic advantage.

CO N FLI C T O F I NTER E S T S TATEM ENT
MA was a co-supervisor of ET and involved in the research about the consumer experience, and was subsequently a participant in the health-care practitioner data set. The remaining authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
ET contributed to the conception, design, acquisition of data, analysis and interpretation of the consumer data. ET was also involved in interpretation of health-care practitioner data compared with the consumer data and drafting this manuscript. ET approves of the ver-

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The study was approved by The University of Melbourne (HREC 1545806.3 and HREC 1646785.9), and informed consent was obtained from all participants. The study was performed in accordance with the Declaration of Helsinki.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on reasonable request from the corresponding author. The data are not publicly available due to privacy/ethical restrictions.