Scientific and ethical challenges in nutrigenomics ...

Scientific and ethical challenges in nutrigenomics & nutrigenetics research A survey of researchers’ perceptionS FINAL REPORT

Omics-Ethics Research Group

–

2013

Published by Omics-Ethics Research Group Bioethics Programmes Department of social and preventive medicine School of Public Health (ESPUM) University of Montreal 7101, Av. du Parc C.P. 6128 Succ. Centre-ville Montreal, QC Canada H3C 3J7 Email: [email protected] Website: www.omics-ethics.org November 2013 © Omics-Ethics Research Group 2013 All rights reserved This report can be viewed or downloaded at www.omics-ethics.org

SCIENTIFIC AND ETHICAL CHALLENGES IN NUTRIGENOMICS & NUTRIGENETICS RESEARCH A SURVEY OF RESEARCHERS’ PERCEPTIONS FINAL REPORT

About The Omics-Ethics Research Group With an empirical ethics program addressing the key ethical and social issues raised by genomics research (and its applications), but also by emerging fields of research such as human enhancement, the OMICS-ETHICS Research Group at the Bioethics Programmes1, University of Montreal performs high-quality research and knowledge translation in an international context. While omics technologies2 are increasingly adopted, the research in this field, as well as its clinical and environmental applications continue to face important ethical and social issues. Socio-ethical engagement with omics technologies, as well as with other emerging biotechnologies, is crucial because the study of ethical aspects could influence and shape the promotion of health and welfare of individuals in a more predictable and sustainable manner while these technologies and associated innovations develop, instead of waiting until the science has already set its course. This has the potential to contribute substantially more to society than what is achievable with post-hoc (i.e., after the fact) ethical inquiries. Omics is a neologism referring to a broad field of inquiry in applied bioscience that employs high throughput biomarker technologies (e.g., genomics, metabolomics), for example, in pharmaceutical research (pharmacogenomics) and nutrition science (nutrigenomics). The overarching goal is to mechanistically analyze the relationships between phenotypic variation (e.g., disease susceptibility, response to drugs and food) and variation in various biological information domains such as in the human genome. With computational biology and bioinformatics tools that allow analyzing and interpreting large-scale omics datasets, omics bioscience research and its attendant social and ethical corollaries gain substantial significance. The OMICS-ETHICS Research Group aims (1) to lay an evidence-based empirical foundation that can both discern and anticipate the socio-ethical issues associated with omics technologies, along with other emerging technologies, such as human enhancement and (2) to study the feasibility of developing tools that can provide appropriate guidance in ethics-related matters to stakeholders in various fields of research and their applications, in particular in the above sub-disciplines as well as in personalized health interventions. For more information, please consult www.omics-ethics.org 1. Bioethics Programme website: http://www.bioethique.umontreal.ca/en/. (Accessed October 2013). 2. For a definition of “omics”, see Omics-Ethics Research Group. What is “Omics” Science? http:// www.omics-ethics.org/en/what-is-omics-science (Accessed October 2013).

iii

Authors of the report Thierry Hurlimann Béatrice Godard The authors wish to thank Nathalie Égalité, Omics-Ethics Research Group, for the grammatical and linguistic proofreading.

Members of the Omics-Ethics Research Group who contributed to the NGx project: Béatrice Godard Director of the Omics-Ethics Research Group Department of social and preventive medicine University of Montreal, Montreal, QC, Canada Janice Graham Technoscience and Regulation Research Unit (TRRU) Dalhousie University, Halifax, N.S., Canada Thierry Hurlimann Department of social and preventive medicine University of Montreal, QC, Canada Darquise Lafrenière Department of social and preventive medicine University of Montreal, QC, Canada Vincent Menuz OMICS-ETHICS Research Group University of Montreal, QC, Canada Julie Robitaille Institute of Nutrition and Functional Foods (INAF) and Department of Food Science and Nutrition Laval University, QC, Canada Raphaëlle Stenne OMICS-ETHICS Research Group University of Montreal, QC, Canada Marie-Claude Vohl Institute of Nutrition and Functional Foods (INAF) and Department of Food Science and Nutrition Laval University, QC, Canada iv

TABLE OF CONTENT Omics-Ethics Research Group ..................................................................................................................... iii Authors and collaborators ........................................................................................................................... iv Acknowledgements ...................................................................................................................................... vi Goals of the report ........................................................................................................................................ 1 CHAPTER 1 – Summary of Phase 1 results and questions raised ............................................................ 3 1.1. Ethnicity of participants in NGx research ......................................................................................... 4 1.2. Location of NGx studies ...................................................................................................................... 5 1.3. Age of participants ............................................................................................................................... 5 1.4. Source of study populations ................................................................................................................ 7 1.5. Potential risks of biohype .................................................................................................................... 7 CHAPTER 2 – Description of Phase 2 – Survey of NGx researchers ...................................................... 11 2.1. Goal and methodology ...................................................................................................................... 12 2.2. Respondents’ characteristics ............................................................................................................. 13 2.3. Limitations .......................................................................................................................................... 14 2.4. Summary of published results .......................................................................................................... 14 CHAPTER 3 – Definitions of nutrigenomics and nutrigenetics ............................................................ 17 CHAPTER 4 – Challenges raised by “ethnicity” in NGx research .......................................................... 21 CHAPTER 5 – Challenges raised by the age of participants in NGx research ....................................... 27 CHAPTER 6 – Ethical issues encountered by NGx researchers .............................................................. 35 Further steps and closing remarks ............................................................................................................ 39 APPENDICES Appendix I: Phase 1: methodology and limitations .......................................................................... 43 Appendix II: Questionnaire ................................................................................................................... 47 Appendix III: Respondents’ comments on the definition of nutrigenomics .................................... 59 Appendix IV: Respondents’ comments on the definition of nutrigenetics ....................................... 65 Appendix V: Respondents’ comments on the statements about ethnicity of participants in NGx research .............................................................................................................................. 71 Appendix VI: Respondents’ comments on the statements about the age of participants in NGx research .............................................................................................................................. 75 Appendix VII: Respondents’ comments on the ethical issues encountered in NGx research .......... 79

v

Acknowledgements The Omics-Ethics Research Group acknowledges funding support from the Fonds de Recherche du Québec – Santé (FRQS) and the Canadian Institutes of Health Research (CIHR).

vi

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions

GOALS OF THE REPORT The present report provides the results of a survey of researchers in nutrigenomics and nutrigenetics [hereafter: “NGx”]. This survey was conducted in 2012 and followed a first phase of a broader research project, A Foundation for Evidence-Based Management of Nutrigenomics Expectations and ELSIs1, aimed at laying an empirical foundation that could discern and anticipate socio-ethical issues associated with NGx research and its applications. In the first phase of this project, an extensive analysis of NGx clinical studies published between 1998 and 2007 inclusively was performed (see Chapter 1). This analysis showed that NGx researchers face specific scientific and ethical issues while conducting and publishing NGx clinical research. The second phase consisted of a survey aimed at exploring their perceptions of such issues. This report focuses on the results of the survey on researchers’ perceptions about the definition of nutrigenomics as opposed to nutrigenetics, as well as the challenges raised by the recruitment of both ethnic minorities and participants of all ages in NGx research. Only results that were not previously published in peer-reviewed journals are presented in this report. Full references of publications linked to this survey and to the research project are provided in Chapter 1 and 2. Results presented in this report will not be discussed, as they closely relate to the issues that were developed in our previous publications, and that are summarized in Chapter 1 and 2. However, we believe that this report may further illustrate current challenges that are faced by NGx researchers, as well as their opinions on such challenges. Our goal is also to share all survey results with the participants in the survey, so as to fulfill our commitment to return our findings to them. The importance of diet to sustain health, prevention and treatment of diseases has been known for a long time. However, the advent of new research tools such as high-throughput “omics” technologies and bioinformatics has recently enabled researchers to go deeper in the analysis of the complex mechanisms that are involved in the way our bodies process and respond to food and thus, ultimately impact human health and well-being. Notably, the knowledge of the human genome has dramatically broadened the scope of studies in nutrition science. NGx is the results of this new alliance between genomics/genetics and nutrition. A strict definition of nutrigenomics and nutrigenetics is difficult, as it often happens with emerging health technologies and new fields of inquiry. A distinction between nutrigenomics and nutrigenetics is often blurred by various and sometimes conflicting definitions, and despite notable differences, both terms are still commonly used interchangeably. For the purposes of this report, however, “NGx” will refer to both nutrigenomics and nutrigenetics, and is restricted to the study of genome/gene(s)-nutrition interactions. NGx aims to understand how genome or/and gene(s)-diet interactions influence individuals’ and populations’ response to food, disease susceptibility, and population health. In this respect, NGx research aims to understand how nutrients and other food components influence genome/gene(s) expression but also how the genome/gene(s) (of individuals or populations) may impact response to diet. Beyond general understanding of the molecular mechanisms involved in nutrition and health, NGx research often tends to focus on diagnostics and knowledge that could be used by subpopulations and/or individuals to customize their diet to prevent disease and promote well-being. 1. A description of the Omics-Ethics group ongoing projects on NGx is available at http://omicsethics.org/en/research-projects-nutrigenomics (Accessed October 2013). © Omics-Ethics Research Group 2013

1

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions While NGx is expected to be implemented in personalized health interventions, a comprehensive inventory of associated ethical-legal-social issues (ELSIs) has yet to be produced. Ethics research on personalized health interventions must take into account the potentially unique ELSIs introduced by different -omics technologies (such as NGx) and the bioscience application context. NGx tests for routine clinical use are at an early stage of development; meanwhile, there are intensive NGx bioscience efforts in both preclinical and clinical research domains. NGx has broad public health significance since it focuses on both patients and healthy populations, in addition to the discernment of nuanced differences in pre-disease states in presently healthy individuals. The strong public health and preventive medicine focus of NGx research as well as the day-to-day importance of food in peoples’ lives may raise ELSIs. Studying both technology-specific ELSIs but also the broader range of ethical issues relevant to personalized health interventions, spanning from clinical discoveries/innovation to publication/translation of clinical -omics research data is crucial because it can influence and positively shape the development of personalized health interventions and help avoid predictable pitfalls, thus ensuring an effective and ethical application of NGx in the laboratory, in the clinic and in the evidence-based development of science policy. The intent of this report and the dissemination of our results is not to argue that NGx and its future applications raise more concerns than other genomics/genetics or nutrition sciences applications, nor to question the validity or relevance of NGx research. The main goals are to raise awareness about potential ethical and scientific pitfalls that could threaten a successful and fair implementation of future NGx applications and to report researchers’ perceptions on such issues.

2

© Omics-Ethics Research Group 2013

CHAPTER 1 SUMMARY OF PHASE 1 RESULTS AND QUESTIONS RAISED



CHAPTER 1 – Summary of Phase 1 results and questions raised In Phase I of this project, an extensive analysis of 173 clinical1 studies published in the field of NGx since 1998 until 2007 inclusively was performed. The methodology used in this first phase can be found in Appendix I. This analysis highlighted both scientific and ethical challenges in NGx research that are summarized on our website2 and addressed in detail in two publications: • Sections 1.1. to 1.4. below are developed in Hurlimann, T., Stenne, R., Menuz, V. and Godard, B. (2011) Inclusion and exclusion in nutrigenetics research: ethical & scientific challenges. Journal of Nutrigenetics and Nutrigenomics 4(6): 322-343; • Section 1.5. below is addressed in Stenne, R., Hurlimann, T., Godard, B. (2012) Are Research Papers Reporting Results from Nutrigenetics Clinical Research a Potential Source of Biohype? Accountability in Research: Policies and Quality Assurance 19(5): 285-307.

1.1. Ethnicity of participants in NGx research Ethnicity matters in NGx research as it is of the utmost importance to consider the influence of acculturation on diet and health, and the impact of genetic admixture in populations and genomic variability on research results3. Yet, our sample of NGx studies provides scant and often ambiguous information about the participants’ race, origin, ancestry, or ethnicity. It appears that there is little coherence between publications in the way the terms race/ethnicity/ancestry/origin and even nationality are used to describe sample populations. Similarly, the way in which ethnicity, race, origin, or ancestry was inferred or measured by researchers is rarely explicitly reported. Information about the participants’ origin could be found in 124 publications (72%), thus it could not be determined for 49 studies (28%) of our sample. Included in the latter category are all publications that did not report any other information about the participants’ origin than the geographical location of the clinical research. Yet, can it be inferred to what populations study findings apply if the ethnicity of participants is not explicitly reported in publications? Is there a way to overcome the controversies and uncertainties linked with the use and the scope of the “race”/ “ethnicity” concepts in NGx research?

1. Based on the NIH definition of clinical research (NIH 2011, Glossary & Acronym. http://grants. nih.gov/grants/glossary.htm. Accessed October 2013), all intervention and/or observational studies that involved human beings as participants were considered, while studies limited to the analysis of human cells or tissues only with no other active human participation were excluded. 2. Omics-Ethics Research Group (2012) Ethics & Nutrigenomics Research: Preliminary results. http://omics-ethics.org/en/NGX-research-index. (Accessed October 2013). 3. Please note that a complete bibliography for the sources consulted to inform/support these statements is provided in the references of the articles listed at the beginning of this chapter.

4


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Our analysis also shows that the majority of participants in our sample of NGx studies are “white”/ Caucasian. In such circumstances, it is crucial to address the issues associated with a lack of representation of ethnic minorities in NGx research. To what extent does this lack of representation limit the possibilities to generalize study findings to ethnic minorities? Is this an issue in terms of justice and equity in research and global health? How should this issue be addressed in light of the many challenges raised by the inclusion of ethnic minorities in NGx research? Among these challenges: • Difficulties in the recruitment and retention of ethnic minorities in clinical research; • Controversies around the use of race and ethnicity as a proxy for genotype; • Socio-ethical issues linked to the potential resurgence of race as a biological concept and the risks of discrimination and stigmatization that study findings may raise; • Adaptation of food frequency questionnaire to diverse cultural dietary habits; • Uncertainty about what constitutes an ‘appropriate’ minority inclusion; • Confusion and ambiguity in the ways ethnicity, race, or ancestry are to be understood, inferred, measured, assessed, and reported; • Impact of the inclusion ethnic minorities on study results due to population stratification and other confounding factors.

1.2. Location of NGx studies Of the 173 publications in our sample of studies, the participants’ location was explicitly reported or could be inferred with certainty as being in Europe for 73 studies (42%) and North America for 68 studies (39%). The other studies were conducted with participants in Asia (17%) and Central America (5%). No NGx study in our sample enrolled participants located in Africa, South America and Oceania. Is this an issue in terms of justice and equity? Could such a finding be the result of the methodological limitations of our study (see Methodology, Appendix I) or of the non-publication of NGx studies conducted in these geographical areas? In any case, in order to achieve its goals and promote optimal health on a global scale, NGx research should also be conducted in developing countries, with local participants and local foodstuffs. Indeed, the usefulness and validity of research results obtained solely in industrialized countries may be limited to these countries.

1.3. Age of participants Eligibility criteria related to the age of participants in our sample of studies were not reported in 67 publications (39%). Our analysis shows that demographic characteristics, such as the age distribution of participants, were often not clearly documented and could not be ascertained in 16% of the published articles in our sample. Yet, can it be inferred to what populations study findings apply if


5

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions eligibility criteria relating to age and the actual age distribution of participants are not clearly reported in publications? Age matters in NGx research as it is of the utmost importance to consider the variations of gene expression and of dietary needs across life span. In particular, much remains to be learnt on elderly-specific nutritional needs and the influence of diet on the aging process. Inadequate nutrition is a major problem affecting elderly people’s health and needs to be further studied. Similarly, several knowledge gaps were identified that pertained to infants’, children’s, and adolescents’ nutritional needs1. However, our results show that middle-aged populations are the most studied populations in NGx research. Less than 10 studies in our whole sample (n = 173) focused on people < 20 years old and no study focused on people ≥ 75 years old. Does such a lack of representation of the elderly and minors in NGx research limit the possibilities to generalize study findings to these populations? Is this an issue in terms of justice and equity?

1. Please note that a complete bibliography for the sources consulted to inform/support these statements is provided in the references of the articles listed at the beginning of this chapter. 6


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Moreover, the elderly (≥75 years old) and people less than 20 years of age were explicitly excluded from a majority of studies in which eligibility criteria relating to age were reported. Yet, is there always a scientific rationale to do so? If not, are such precise age cut-offs arbitrary and thus unethical? Finally, how should such issues be addressed in light of the many challenges raised by the inclusion of the elderly and minors in NGx research? Among these challenges: • Obtainment and quality of informed consent (and assent) in clinical research; • Controversies around the notion of minimal risk in research involving vulnerable people; • Ethical concerns raised by genetic testing on minors; • Validity and reliability of tools used to assess nutritional intake or behaviors in older participants, infants, children, and adolescents.

1.4. Source of study populations In 59% of studies in our sample, the authors reported that participants were either part of, or were recruited from, previous or ongoing (usually larger) studies. As we found that such information was not always explicitly reported in the publications of our sample, the number of studies in which pre-existing cohorts were used is actually higher. It must be noted that several studies (addressing different NGx interactions) in our sample were conducted on the same populations or on participants who were recruited from or part of the same pre-existing cohorts. As mentioned above, our results show that in our sample, NGx interactions are mostly described for ‘‘white’’/Caucasian populations, located in North America and Europe, while from a public and global health perspective, appropriate inclusion and representativeness of all groups and populations in NGx research is of the utmost importance. This may raise concerns: could the recruitment from pre-existing cohorts in order to test new hypotheses in NGx reinforce this trend and limit the application of research findings to populations whose demographics do not diverge substantially from that of these pre-existing research samples?

1.5. Potential risks of biohype Biohype occurs when the anticipated or claimed future benefits and promises of a developing field of science turn out to be overstated, exaggerated, or premature, with regards to the data and knowledge that have been gathered. NGx may be very vulnerable to biohype, because: • There are great expectations associated with the prevention and the treatment of many diseases targeted by NGx future applications (e.g., cancer, diabetes, cardio-vascular diseases); • The potential market for NGx applications is quite vast. Potential NGx applications may apply to both sick and healthy individuals, for disease prevention as well as well-being promotion. In addition, food intake is an essential daily activity; there are thus numerous opportunities for NGx applications.


7

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions • NGx is a complex research area with methodological limitations1 that may blur the scope or clinical significance of study results, and thus limit their extrapolation to clinical applications. The media have been identified as one of the main channels of biohype for the general public because they may misinterpret the data that are reported in scientific literature, and thus disseminate premature or exaggerated claims about the benefits or clinical significance of research findings2. One of the goals of the first phase of our project was to investigate whether scientific articles could contribute to feeding a biohype phenomenon. Our analysis showed that a minority (9%) of the publications in our sample contained explicit claims about broader implications of research findings, i.e., clinical applications extrapolated from study results, despite significant methodological limitations. Moreover, all these limitations were not always reported in these publications. While these results show that published NGx clinical studies do not appear as a primary source of biohype, there are crucial questions that must be addressed: • Does the complexity of NGx research and the limitations resulting from the methodological tools used in such research (e.g., questionnaires, study design, etc.) raise concerns about the extrapolation of study results into specific clinical recommendations? Do such limitations raise concerns about the achievability of the potential benefits of NGx? • Could the extrapolation of study results and the reporting of potential benefits associated with the development of NGx in scientific publications generate a risk of biohype, in particular if all methodological limitations are not explicitly acknowledged? • To prevent such a risk, should researchers avoid extrapolating their results into clinical applications whose achievability and efficacy is still uncertain? Should they explicitly report all study limitations and discuss their impact on the interpretation of study findings to prevent any misinterpretation on the part of readers? • Can expected benefits of NGx be claimed without also considering the potential socio-ethical implications of NGx applications in public health? For instance, could potential applications of NGx raise fears of discrimination or stigmatization, similarly to other genetic tests? What impact could dietary recommendations have on the socio-cultural meanings of food? Is there a risk of medicalization of diet?

1. These methodological limitations are extensively discussed in Stenne, R., Hurlimann, T., Godard, B. (2012) Are Research Papers Reporting Results from Nutrigenetics Clinical Research a Potential Source of Biohype? Accountability in Research: Policies and Quality Assurance 19(5): 285-307. 2. Ibid.

8




9

CHAPTER 2 DESCRIPTION OF PHASE 2 SURVEY OF NGx RESEARCHERS


CHAPTER 2 – Description of Phase 2 – Survey of NGx researchers 2.1. Goal and methodology To ensure an in-depth interpretation of Phase 1 results, and address the questions raised by these results, we drew upon the expertise of researchers at the forefront of NGx development and designed an English electronic survey in which NGx researchers were asked to share their perceptions about the challenges and issues identified in Phase 1. The questionnaire contained 64 items and was divided in 6 main parts: a) Definition of nutrigenomics as opposed to nutrigenetics; b) General assertions about the risks and benefits of NGx; c) Time envisioned for the achievement of benefits; d) Scientific and ethical challenges pertaining to NGx clinical research (in particular to participants’ ethnicity and age); e) Content of scientific publications reporting results of NGx clinical studies and audiences targeted by such publications; f) Demographics. The questionnaire can be found in the Appendix II. The whole questionnaire, which was pre-tested among a small group of NGx researchers, took approximately 20 minutes to complete. Ethical approval was obtained from the Research Ethics Board of the University of Montreal. It was presumed that consent to participate in the survey was given by completing and submitting the questionnaire. SOM Inc. (Quebec) conducted the online survey. From January 17th to February 25, 2012, 586 researchers were sent a personalized e-mail invitation to participate in the survey. An email reminder was sent to those who had not responded one week after the original invitation, with a follow-up reminder 11 days later. The 586 NGx researchers who were invited to participate in our survey were: a) First or corresponding authors in NGx clinical studies published between 1998 to 2011 inclusively; b) Members of societies, associations and other research groups identified as NGx experts through an Internet Google search using the keywords “nutrigenetics” or “nutrigenomics” (e.g., International Society of Nutrigenetics and Nutrigenomics, NuGo, Nutrigenomics Consortium Netherlands, Nutrigenomik Deutschland), or their own website; c) Selected among the speakers of the 4th Congress of the International Society of Nutrigenetics and Nutrigenomics; or d) First or corresponding authors of articles on gene and food/ nutrient interactions published in the Journal of Nutrigenetics and Nutrigenomics; or e) Corresponding authors of articles in the journals Gene & Nutrition and Molecular Nutrition and Food Research, identified through a search with the keywords “nutrigenetics” or “nutrigenomics” within these 2 journals. Cross tabulation χ2 analysis allowed us to draw comparisons between respondents according to their answers and characteristics (e.g. demographics). We considered a P-value of 0.05 or less as statistically significant.

12



2.2. Respondents’ characteristics In all, 126 online questionnaires were completed, resulting in a response rate of 22%.

Respondent characteristic Gender Male Female Age ≤30 31-40 41-50 51-60 ≥61 Continent of residence Europe1 North America2

%

Other3 Professional status Professor4 Other5

11

Not specified Affiliation Private entity (part or full)6

1

Public entity only7 Experience in research (years)8 < 10 10-19 20-29 ≥ 30 Diseases targeted by respondents’ research Metabolic diseases (e.g., diabetes, obesity) Cancer HBVD9 Combination of the preceding categories Other10 Not specified Respondents’ research conducted in:11 Europe North America Other12 Not specified Sources of funding for respondents’ research: Public only13 Private (part or full)14 Not specified

88

Notes

52 48 3 27 29 28 13 58 31

73 26

12

17 33 27 23

23 15 9 9 27 17 64 43 13 2

Europe: UK, Netherlands, Spain, Italy, Germany, France, Sweden, Denmark, Finland, Greece, Switzerland, Croatia, Austria, Czech Republic, Norway, Poland 2 North America: United States, Canada 3 Other: Brazil, China, Malaysia, New Zealand, India, Japan, Australia, Israel, Singapore 4 “Professor” includes full, associate and assistant professors. 5 Other: respondents who did not indicate “professor”, and included at least one of the following: MD, post-doctoral student, research director, institute/contract research organization director/president, senior scientist, dietician, researcher, research fellow, retired, dean, PhD student, plastic surgeon, science communication professional, chief operating officer 6 “Private entity” includes: private laboratory or organization, private hospital/clinic 7 “Public entity” includes: university, public hospital/clinic, public research institute, non-profit organization (e.g., trust, charity, research organization), government science laboratory 1

8

51% of the respondents have done research for at least 20 years

Heart and Blood Vessel Diseases (HBVD) Other diseases or fields of research such as autoimmune diseases, molecular nutrition, foetal development, aging, gut microbial gene and physiology, food safety, systems biology, basic research in various fields, etc. 9

10

Total > 100 as some respondents conduct research in > 1 continent Other: Australasia (Australia and New Zealand), South America (Brazil), Asia (China, Malaysia, India, Japan, Singapore, Israel), Africa (Morocco) 11 12

Funding from government agencies, universities, not-for-profit organisations (trusts, charities, foundations, etc.), European community, including mentions of “European project” and “International funding” 14 Were considered as sources of “private funding: funding originating from for-profit companies and/or private/public partnerships 13

64 30 6


13


2.3. Limitations NGx research is an interdisciplinary field, with borders blurred by the complex processes under study. Thus, it is likely that our search and methodology to identify researchers in this field overlooked some, particularly in non-Western countries. Perceptions of NGx researchers about the issues addressed in our questionnaire may differ across countries. Moreover, while respondents were geographically dispersed, the survey was only carried out in English, which may have had an impact on participation rate.

2.4. Summary of published results Two peer-reviewed publications report the results of the survey regarding researchers’ perceptions about potential benefits and risks raised by NGx and its future applications.

2.4.1. Benefits associated with nutrigenomics research and their reporting in the scientific literature: researchers’ perspectives This paper by Stenne, R., Hurlimann, T., and Godard, B., published in Accountability in Research: Policies and Quality Assurance 2013; 20(3): 167-183, focuses on the potential benefits of NGx and the time envisioned for their realization. In the survey, respondents had to assess five statements describing expected benefits associated with the development of NGx. The following five statements were extracted from scientific, peer-reviewed articles reporting NGx clinical research results: • “More effective preventive and therapeutic interventions will be developed within the context of nutritional genomics as new tools become available to achieve effective dietary prevention and therapy.” • “More precise public health advice about dietary intake, supplement use, and genetic testing will be possible through nutrigenomics research.” • “Personalized dietary recommendations will become more effective through the incorporation of genetic information.” • “Nutrigenomics will allow the identification of individuals who suffer from adverse reactions caused by the consumption of certain nutrients.” • “Dietary interventions based on genotypic knowledge will be introduced as a useful strategy for the prevention of some complex diseases.”

14


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions A vast majority of respondents are in agreement with these benefits (between 68% and 83%, depending on the benefit). However, researchers (between 36% and 60%) have serious doubts on the achievability of these benefits within the next five years (as of February 2012). The results of the survey also show that NGx researchers have different and even opposing views about (i) the impact of the methodological limitations on the achievability of benefits and (ii) the uncertainty about the longterm efficacy of future NGx interventions. This first paper also addresses the motives that may lead NGx researchers to explicitly mention potential benefits in scientific peer-reviewed articles and the audience(s) they target when reporting their results in such articles. Researchers are almost unanimous (90%) on the importance of discussing the consequences of methodological limitations on the achievability of NGx potential benefits in scientific articles. However, they are divided when it comes to the relevance of mentioning benefits whose achievability is uncertain in scientific publications. A majority of respondents tend to agree that mentioning potential benefits of applications that could result from study findings i) is useful to draw the readers’ attention to the advantages of the development of NGx, ii) helps foster continuing funding of NGx research and iii) gives the manuscript a better chance of being accepted for publication. Finally, the results of the survey show that when reporting NGx clinical research results in a scientific article, almost all researchers include peers as a target audience. These are considered as the first target audience by a vast majority of researchers (88%). The second most targeted audience, or the first audience outside the research community, is health service providers (95%). Policy makers, the general public, and the media are considered less crucial audiences, but are nonetheless among the target audiences for a majority of researchers (ranging from 62% to 77%). Detailed results and discussion can be found in the full article.

2.4.2. Risks of nutrigenomics and nutrigenetics? What the scientists say This second paper, by Hurlimann, T., Menuz, V., Graham, J., Robitaille, J., Vohl, M.C., and Godard, B., accepted for publication in Genes & Nutrition1, focuses on the opinions of NGx researchers about potential risks raised by the implementation of NGx applications. The results of the survey show that most researchers downplay these potential risks. They do not believe that NGx will reconfigure foods as medication or transform the conception of eating into a health hazard. The majority think that NGx will produce no added burden on individuals to get tested or to remain compliant with NGx recommendations, nor that NGx will threaten individual autonomy in daily food choice. Finally, the majority of researchers do not think that NGx will lead to discrimination against and/or stigmatization of people who do not comply with NGx dietary recommendations. Despite this optimism among NGx researchers, we suggest that the potential benefits of NGx, along with other genetic or genomic clinical applications, require balancing with potential socio-ethical risks. In particular, tension between socio-cultural meanings surrounding the medicalization of food, and the impact of NGx on individual autonomy and responsibility, warrant further attention and are addressed in this article. 1. The manuscript was accepted for publication in October 2013, and will be published in Genes & Nutrition in January 2014. © Omics-Ethics Research Group 2013

15

CHAPTER 3 DEFINITIONS OF NUTRIGENOMICS AND NUTRIGENETICS


CHAPTER 3 – Definitions of nutrigenomics and nutrigenetics NGx researchers were surveyed concerning their views of the following definitions of nutrigenomics and nutrigenetics: “Nutrigenomics studies the functional interactions of food and its components with the genome at the molecular and cellular level, as well as the systemic level, with the goal of using diet to prevent or treat diseases.” “Nutrigenetics examines the effect of genetic variation on the interaction between diet and disease. Its goal is to generate recommendations regarding the risks and benefits of specific diets or dietary components to the individual.” The definition of nutrigenetics was taken from Ordovas and Mooser’s article “Nutrigenomics and nutrigenetics” published in Curr. Opin. Lipidol. 2004; 15: 101-108. The definition of nutrigenomics was adapted from the same article. Answer options were: “Yes, totally”, “Yes, partly” and “No”. Respondents were given the option to provide comments to explain their answer. A large majority of NGx researchers (96%) totally or partly agreed with the proposed definition of nutrigenomics (Figure 1). Figure 1 – Definition of nutrigenomics

Figure 1: Researchers’ agreement (n = 126) with the definition of nutrigenomics. The question read as follows: “Do you agree with the following definition of nutrigenomics?: ‘Nutrigenomics studies the functional interactions of food and its components with the genome at the molecular and cellular level, as well as the systemic level, with the goal of using diet to prevent or treat diseases.’ (Curr Opin Lipidol. 15(2); 2004)”.

18


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Forty-two researchers (33%) provided comments to explain their answer (Appendix III). Most of those who commented partly agreed or disagreed with the proposed definition and expressed concerns or doubts about the goal of nutrigenomics as defined (i.e. “using diet to prevent or treat diseases”). They thought that the goal of nutrigenomics should be broadened and should not be limited to the prevention or treatment of diseases. According to these respondents, it should include the understanding of genome-food interactions in general, as well as health optimization, with no necessary focus on disease prevention. Several researchers explicitly stated that the definition of nutrigenomics should not include any goal more specific than the “study of the effects of nutrients on gene expression”. All comments (n = 31) describing doubts or concerns about the goal of nutrigenomics as outlined in the definition, or proposing a definition that does not include a goal of this sort, can be found in yellow cells in Appendix III. Most respondents (94%) agreed or partly agreed with the proposed definition of nutrigenetics (Figure 2). Forty-three researchers (34%) provided comments to explain their answer (see Appendix IV). A majority of these respondents expressed doubts and concerns that were similar to the ones expressed about the goal of nutrigenomics as reported above (i.e. goal should be broadened or not be mentioned at all in the definition). In addition, three respondents stressed that the goals of nutrigenetics should not or could not be targeting individuals (or targeting them exclusively), but rather sub-groups in the population (or sub-groups in addition to individuals) (Appendix IV). A few others mentionned that “generating recommendations’’ was a “premature” or an “unreachable” goal for many diseases. All comments can be found in Appendix IV. Figure 2 – Definition of nutrigenetics

Figure 2: Researchers’ agreement (n = 126) with the definition of nutrigenetics. The question read as follows: “Do you agree with the following definition of nutrigenetics?: ‘Nutrigenetics examines the effect of genetic variation on the interaction between diet and disease. Its goal is to generate recommendations regarding the risks and benefits of specific diets or dietary components to the individual.’ (Curr Opin Lipidol. 15(2); 2004)”.


19

CHAPTER 4 CHALLENGES RAISED BY ‘‘ETHNICITY’’ IN NGx RESEARCH


CHAPTER 4 – Challenges raised by ‘‘ethnicity’’ in NGx research Respondents first had to indicate their level of agreement (“Strongly agree”, “Generally agree”, “Somewhat disagree” and “Strongly disagree”) with the following statements: “A lack of representation of ethnic groups in NGx research limits the possibilities of generalizing the results to these populations” (Figure 3) “Ethnicity should be reported in scientific publications of NGx research, otherwise it cannot be inferred to what populations study findings can apply” (Figure 4) “The concepts of ethnicity, race or origin are vague and controversial. Thus, it is difficult to use them in NGx scientific publications” (Figure 5) Respondents were given the option to provide comments, if any, after answering the aforementioned questions. NGx researchers were also asked what challenges, if any, they perceived in the recruitment and participation of ethnic minorities in NGx research. Twelve methodological or ethical challenges were proposed (see Figure 6). A category “Other” allowed respondents to suggest other challenges. Respondents could choose as many answers as they wanted. A majority of the respondents (79%) strongly or generally agreed with the idea that a lack of representation of ethnic groups in NGx clinical research limits the possibilities of generalizing the research results to these populations (Figure 3). Figure 3 - A lack of representation of ethnic groups in NGx research limits the possibilities of generalizing the results to these populations

Figure 3: Researchers’ agreement (n = 126) with the impact of a lack of representation of ethnic groups in NGx research. The question read as follows: “Please indicate your level of agreement on the following statement about ethnicity in NGx clinical research: ‘A lack of representation of ethnic groups in NGx research limits the possibilities of generalizing the results to these populations’.”

22


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Similarly, a large majority (94%) believed that it cannot be inferred to what ethnic groups research findings may apply if ethnicity is not reported in scientific publications (Figure 4). None of the respondents who somewhat or strongly disagreed with this statement (6%; Figure 4) explicitly commented on it. Figure 4 - Ethnicity should be reported in scientific publications of NGx research, otherwise it cannot be inferred to what populations study findings can apply

Figure 4: Researchers’ agreement (n = 126) with the impact of not reporting ethnicity in NGx scientific publications. The question read as follows: “Please indicate your level of agreement on the following statement about ethnicity in NGx research: ‘Ethnicity should be reported in scientific publications of NGx research, otherwise it cannot be inferred to what populations study findings can apply’.”


23

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions A majority of the respondents (73%) disagreed with the statement “The concepts of ethnicity, race or origin are vague and controversial. Thus, it is difficult to use them in NGx scientific publications” (Figure 5). Several researchers explicitly commented on this statement (see Appendix V). In short, a majority of those who provided comments stressed that, whether vague and controversial or not, ethnicity had to be reported in publications. One respondent suggested using the concept of “ethnogeographic origin” and one respondent explicitly stated that there existed “accepted terms for ethnicity/ race/ etc based on Census data” (Appendix V). Figure 5 - The concepts of ethnicity, race or origin are vague and controversial. Thus, it is difficult to use them in NGx scientific publications

Figure 5: Researchers’ agreement (n = 126) with the impact of controversies about “race” and “ethnicity” on the use of such concepts in scientific publications. The question read as follows: “Please indicate your level of agreement on the following statement about ethnicity in NGx clinical research: ‘The concepts of ethnicity, race or origin are vague and controversial. Thus, it is difficult to use them in NGx scientific publications’.” The results of the survey show that 84% of the respondents perceived at least one challenge in the recruitment and participation of ethnic groups in NGx research (Figure 6). Nearly half of researchers (48%) anticipate cultural issues when recruiting ethnic minorities, and about one third (36%) thought that members of ethnic minorities might not be willing to participate in research. Controversies about the use of the concepts of “race”, “ethnicity” or “ancestry” in biomedical research, language, risks of stigmatisation or discrimination, and informed consent were deemed less likely, but nonetheless possible challenges (between 13% and 20%) (Figure 6). Methodological/scientific challenges are also cited: measurement/assessment of ethnicity (47%), adaptation of questionnaires to ethnic particularities (38%), measurement/assessment of dietary intake (33%), impact of population stratification on study results (30%), compliance (20%) and retention of participants (16%) (Figure 6). Our analysis shows that 73% of the respondents chose one or more of these methodological/scientific challenges (data not shown in Figure 6). Some respondents (9%) mentioned other challenges such as: difficulties reaching a statistical significant sample size, time needed to overcome all other challenges, subgroup analyses, added confounding factors, genetic background, and variations in dietary habits and food preparation. One researcher stressed that the

24


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions same challenges were present when recruiting Caucasians. Another mentioned having conducted research involving ethnic minorities with a high willingness to participate, who complied with study procedure and who did not withdraw from the studies. One respondent concluded that questions about ethnicity in the survey indicated an overreaction on this matter. Finally, it is worth noting that when asked whether he or she had ever encountered ethical issues while conducting NGx research (see Chapter 6 below), one respondent explicitly mentioned having had trouble in defining “Caucasian” and justifying the exclusion of other ethnic groups from his or her protocol to the Research Ethics Board (see Appendix VI, Category 4). Figure 6 - Challenges in the recruitment and participation of ethnic minorities in NGx research

Figure 6: Researchers’ agreement (n = 126) with challenges in the recruitment and participation of ethnic minorities in NGx research. The question read as follows: “What challenges, if any, do you see in the recruitment and the participation of ethnic minorities in NGx research? (You may choose several answers).”


25

CHAPTER 5 CHALLENGES RAISED BY THE AGE OF PARTICIPANTS IN NGx RESEARCH


CHAPTER 5 – Challenges raised by the age of participants in NGx research Respondents first had to indicate their level of agreement (“Strongly agree”, “Generally agree”, “Somewhat disagree” and “Strongly disagree”) with the following statements: “A lack of representation of infants, children and minors in NGx research limits the possibilities of generalizing the results to these populations”. (Figure 7) “A lack of representation of the elderly in NGx research limits the possibilities of generalizing the results to these populations”. (Figure 9) “There is no scientific rationale to include the elderly (75 years old and over) in NGx research”. (Figure 10) “If the age of participants is not reported in scientific publications of NGx research, it cannot be inferred to what populations study findings apply”. (Figure 12) Respondents were offered the option to provide comments after answering the aforementioned questions. NGx researchers were also asked what challenges, if any, they perceived in the recruitment and participation of infants, children and minors, as well as of the elderly in NGx research. Various methodological or ethical challenges were proposed (see Figures 8 and 11). A category “Other” allowed respondents to suggest other challenges. Respondents could choose as many answers as they wanted.

28


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions A majority of the respondents (73%) strongly or generally agreed with the idea that a lack of representation of infants, children and minors in NGx clinical research limits the possibilities of generalizing the research results to these populations (Figure 7). Figure 7 - A lack of representation of infants, children and minors in NGx research limits the possibilities of generalizing the results to these populations

Figure 7: Researchers’ agreement (n = 126) with the impact of a lack of representation of infants, children and minors in NGx research. The question read as follows: “Please indicate your level of agreement on the following statement about participants’ age in NGx clinical research: ‘A lack of representation of infants, children and minors in NGx research limits the possibilities of generalizing the results to these populations’.”

Yet, most researchers acknowledged methodological and/or ethical challenges when recruiting minors for research (Figure 8). More than a half of them (58%) perceived parental involvement as challenging (Figure 8). A similar proportion of them (54%) agreed that controversial ethical requirements for minimal risk or acceptable burden in research involving minors are problematic (Figure 8). Ethical concerns about genetic testing in children were also cited as an issue by 46% of the respondents (Figure 8). Methodological challenges such as assessment and measurement of dietary intake in infants and other minors, or adaptation of questionnaires were mentioned by 39% and 26% of respondents, respectively (Figure 8). Some respondents (9%) mentioned other challenges such as: difficulties in obtaining Research Ethics Boards’ (REBs) approval for research involving children in general, including in fields other than NGx; parental resistance, bias due to parental involvement, “loss of liberty of behaviour” for children enrolled in research, and consideration of the impact of dietary intake during mother’s pregnancy. Some respondent stressed that a lack of representation of children or challenges with their recruitment and participation in research were not specific to NGx.


29

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Figure 8 - Challenges in the recruitment and participation of infants, children and minors in NGx research

Figure 8: Researchers’ agreement (n = 126) with challenges in the recruitment and participation of infants, children and minors in NGx research. The question read as follows: “What challenges, if any, do you see in the recruitment and the participation of infants, children and minors in NGx research? (You may choose several answers).”

In line with the aforementioned results, a majority of respondents (66%) – although a smaller percentage than for the case of infants, children and minors previoulsy described – were convinced that a lack of representation of the elderly in NGx clinical research might limit the possibilities to generalize research findings to this population (Figure 9). Moreover, an overwhelming majority (93%) thought that the inclusion of the elderly in NGx research is scientifically justified (Figure 10).

30



Figure 9 - A lack of representation of the elderly in NGx research limits the possibilities of generalizing the results to these populations

Figure 9: Researchers’ agreement (n = 126) with the impact of a lack of representation of the elderly in NGx research. The question read as follows: “Please indicate your level of agreement on the following statement about participants’ age in NGx clinical research: ‘A lack of representation of the elderly in NGx research limits the possibilities of generalizing the results to these populations’.”

Figure 10 - There is no scientific rationale to include the elderly (75 years old and over) in NGx research

Figure 10: Researchers’ agreement (n= 126) with the existence of a scientific rationale to include the elderly in NGx research. The question read as follows: “Please indicate your level of agreement on the following statement about participants’ age in NGx clinical research: ‘There is no scientific rationale to include the elderly (75 years old and over) in NGx research’.”


31


However, a substantial majority of researchers (80%, data not shown in Figure 11) foresaw at least one of the following methodological/scientific challenges when conducting research with the elderly: pre-existing medication regimen (67%), comorbidities (62%), measurement and assessment of dietary intake (30%), compliance (20%), retention of participants (19%) and adaptation of questionnaires (13%) (Figure 11). Moreover, around one third of respondents consider that the involvement of relatives or surrogates, the absence of willingness to participate, as well as controversial ethical requirements for minimal risk or acceptable burden in research involving vulnerable people presented challenges (Figure 11). Obtaining informed consent, protecting participants’ autonomy and risks of stigmatisation or discrimination were deemed less problematic (between 9% and 20%) (Figure 11). Some respondents (6%) mentioned other challenges such as: limited time-scale for follow-up and prospective observation due to health/cognition deterioration, resistance on the part of researchers, and resistance on the part of Research Ethics Boards. Figure 11 - Challenges in the recruitment and participation of the elderly in NGx research

Figure 11: Researchers’ agreement (n = 126) with challenges in the recruitment and participation of the elderly in NGx research. The question read as follows: “What challenges, if any, do you see in the recruitment and the participation of the elderly in NGx research? (You may choose several answers).” 32



Figure 12 - If the age of participants is not reported in scientific publications of NGx research, it cannot be inferred to what populations study findings apply’

Figure 12: Researchers’ agreement (n = 126) with the impact of not reporting age in NGx scientific publications. The question read as follows: “Please indicate your level of agreement on the following statement about ethnicity in NGx research: ‘If the age of participants is not reported in scientific publications of NGx research, it cannot be inferred to what populations study findings apply’.” Most researchers (82%) agreed on the necessity of reporting the age of participants in NGx publications in order to clearly identify to what populations study findings can apply (Figure 12). Finally, fifteen respondents (12%) provided comments after answering the questions about the age of participants in NGx research. In most comments, respondents acknowledged that age was an important factor in NGx research, much like in the study of genetics/genomics and nutrition sciences in general. All comments can be found in Appendix VI.


33

CHAPTER 6 ETHICAL ISSUES ENCOUNTERED BY NGx RESEARCHERS


CHAPTER 6 – Ethical issues encountered by NGx researchers Nearly a quarter of respondents (23%) reported having encountered some ethical issues while conducting NGx research (Figure 13). Among them, two respondents did not comment or explain their answer. The comments and explanations provided by all other respondents (n = 27) can be found in Appendix VII. We classified respondents’ comments into five categories, as shown in Table 1 below. Figure 13 - Have you ever encountered any ethical issues in conducting your NGx projects?

Figure 13: NGx researchers’ (n = 126) answer to the question: “Have you ever encountered any ethical issues in conducting your NGx projects?’’

The majority of the respondents who provided comments expressed concerns about the disclosure of individual research results or incidental findings to participants (Appendix VII, Category 1). It appears that determining what individual results should be returned to participants is ethically challenging, but also challenging on practical and scientific terms (e.g., what is a “meaningful” or clinically “significant” result? What if an individual result is deemed clinically significant, but participants previously consented not to be informed?). Five respondents indicated that they had faced issues in the informed consent process (for instance consent for the use of human samples, consent and return of individual results, see Appendix VII, Category 2). Four respondents mentioned difficulties in getting ethical approval from Research Ethics Borads (regarding, for instance, storage and use of human samples in research, definition of “Caucasian” and exclusion of other ethnic groups from research protocol; see Appendix VI, Category 3). Three respondents faced ethical issues linked with ethnicity (such as getting ethical approval for the exclusion of other ethnic groups than Caucasians in a research protocol, and more generally, coping with genetic variations in different populations, see Appendix VI, Category 4). Finally, three respondents mentioned other issues (namely population scepticism “against everything associated with genes”, compliance by participants, and unjustified or unnecessary suffering to animal disease models), while one respondent stressed that all animal and human studies raise ethical issues.

36


Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Table 1 Category number 1. 2. 3. 4. 5.

Ethical issues Issues linked to the disclosure of individual incidental and/or research results (including disclosure to family members) and more generally to the identification of genetically at risk participants in NGx research Issues linked to informed consent Difficulties to get ethical approval or issues raised by Research Ethics Boards Issues linked to ethnicity or genetic variations across populations Others

Number of comments 13 5 5 3 4

Table 1: Researchers’ comments on ethical issues they encountered while conducting NGx research. Twenty-seven respondents commented their answer (out of 29 respondents who stated having encountered some ethical issues). When appropriate, certain comments were split into specific parts that could be classified into different categories. The same comment could also appear in more than one category. Thus the total number of comments in this Table is > 27. All comments can be found in Appendix VII.


37

FURTHER STEPS AND CLOSING REMARKS


Further steps and closing remarks NGx moves across nutrition sciences, genomics/genetics, and public health, with diverse stakeholders involved in its development and implementation, as well as in its funding. All stakeholders’ expertise and collaboration are needed to determine how the challenges linked to the inclusion of ethnic minorities and people of all ages worldwide will be engaged. NGx researchers should not bear the responsibility of engaging such issues alone. In practice, there are certainly scientific reasons that may justify exclusion of certain groups of the population in specific protocols (eg., disease studied, limited ethnic representation in specific geographical locations, sample size issues, etc.) Yet, a substantial majority of surveyed NGx researchers acknowledge that a lack of representation of certain groups of the population in NGx research may limit the possibilities of generalizing the results to these groups. In terms of global health, both industrialized and emerging countries are facing a growing epidemic of the same chronic diseases as those that are the most often studied in NGx research and both have to cope with significant inequalities in health care access1. In such a context, NGx call for the right of whole populations to benefit from its results. Further steps to promote NGx research that could benefit all communities and populations could be: • A clear recognition from all stakeholders involved in the development of NGx and personalized dietary interventions that a lack of representation of ethnic minorities and people of all ages worldwide limit the possibilities of generalizing the results to these groups; • A clear recognition from all stakeholders involved in the development of NGx and personalized dietary interventions that knowledge from NGx research can also be a powerful tool to understand the long-term consequences of underfeeding and malnutrition and could be used to improve prevention and nutrition programs targeting whole populations, including in developing countries; • The inclusion of socio-economic-environment considerations in the initial design of research protocols, with the aim of anticipating potential limitations in the implementation of NGx future applications in various populations; • A clear recognition of the impact of study limitations on the external validity of results;

1. Godard B, Hurlimann T. Nutrigenomics for global health: ethical challenges for underserved populations. Curr Pharmacogenomics Person Med 2009; 7: 205–214.

40



• A consensus on the most effective ways to report participants’ race, ethnicity, ancestry, or origin; • A critical, but also pragmatic, ethical review of inclusion and exclusion criteria in NGx research protocols, with a full recognition from REBs that practical limitations may impede an “ethically ideal population representation” in the specific research protocols they review.


41

APPENDIX I PHASE 1: METHODOLOGY AND LIMITATIONS

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions APPENDIX I A FOUNDATION FOR EVIDENCE-BASED MANAGEMENT OF NUTRIGENOMICS EXPECTATIONS AND ELSIS Phase 1 - Methodology A Foundation For Evidence-Based Management Of Nutrigenomics Expectations And ELSIs is a project by the OMICS-ETHICS research group aiming at laying an empirical foundation that could discern and anticipate the socio-ethical issues associated with NGx research and its potential applications. For more details and contacts, please visit our website at http://www.omics-ethics.org/en/research-projects-nutrigenomics. Methodology of the first phase: literature review, data extraction and analysis A. Sample and analysis We performed a systematic review of clinical studies in nutrigenetics and nutrigenomics published since 1998 until 2007 inclusively. Studies for potential inclusion were identified through a PubMed search. As nutrigenomics is a recent word that was indexed and introduced as a MeSH category encompassing nutrigenetics in PubMed only as of 2008, we completed our search with different combinations of the following key words: ‘food’, ‘nutrient’, ‘diet’, ‘gene’, ‘interaction’, and ‘association’. The search was limited to titles and abstracts of original articles and reviews; commentaries and letters were excluded. Based on the NIH definition [1], all intervention and/or observational studies that involved human beings as participants were considered as clinical studies, while studies limited to the analysis of human cells or tissues only, with no other active human participation than tissue or cell donation, were excluded. Moreover, to be included in our sample, studies had to meet the following criteria: (1) having a specific gene component (e.g. candidate gene or polymorphism); (2) having a dietary component: nutrient (from the diet or as a supplement), food, dietary pattern (e.g. Mediterranean diet), etc., and (3) describing an interaction between (1) and (2) that may impact health, disease onset, or nutritional biological pathways. A total of 173 studies met these criteria. These publications constituted the sample upon which we performed a detailed content analysis. The data extracted from these publications for this first phase of this project were: (a) the authors’ geographical location; (b) the participants’ geographical location; (c) the participants’ particulars, such as race, ethnicity, origin, nationality, ancestry, age, sex, comorbidities, and any other available data linked to participants’ description, as well as any exclusion and inclusion criteria reported by the authors. The description of participants was completed in many cases with the tables provided by the authors to present their results. The following elements were also extracted from the publications of our sample: (d) all authors’ statements or comments about the potential or actual impact of genetic variations linked to ethnicity, and (e) all limitations of the study results explicitly acknowledged and reported by the authors. It has to be noted that in 120 of the 173 publications of our sample (hereafter ‘referring publications’), the authors referred to previous publications and/or previous or ongoing studies (hereafter ‘referred publications’) for the description of the methodology of their study, and/or of their participants or of their selection. Consequently, we consulted all referred publications (190 publications) to gather as much information as possible about the participants’ location, age, and ethnic origin.

44


As the goal of this first phase of our study was also to investigate whether scientific articles could be a source of biohype, we used a method that involves the study of the context in which claims, especially interpretative claims, are made in the scientific literature. An interpretative claim occurs when authors “refer to the broader implications of research findings . . .” [2, p. 11]. Ours goals were to examine i) whether such claims could be found in the articles reporting nutrigenetics study results, and ii) whether articles containing such claims reported the methodological limitations of the study and explicitly acknowledged the impact of these limitations when extrapolating the results to clinical applications. In order to identify any claims that could make scientific journal articles a potential source of biohype and document their prevalence, the “main statements about the utility and scope of the study results” were extracted from each article. These statements were usually found either in the discussion or in the conclusion section of the articles. The “main statements about the utility or the scope of the results of the study” such as reported in the 173 articles were assessed to identify interpretative claims, i.e. explicit statements about potential clinical applications that were extrapolated from the study results. In order to evaluate the extent of the gap between extrapolated clinical applications and their actual achievability given the methodological limitations involved in any given study, the following data were extracted from the articles: a) the gene and polymorphism under study; b) the food component under study; c) the study design; d) the tools used for the assessment of dietary intakes; e) the size of the studied population; f) the impact of genetic variations under study; g) the main statements about utility and scope of the study results; and h) all limitations of the study results explicitly reported and, if it is the case, acknowledged by the authors. A detailed content analysis was subsequently performed.

Appendix I


B. Limitations Given the recent use of the terms ‘nutrigenomics’ and ‘nutrigenetics’, the results of our search for clinical studies in this field is likely limited by the key words (and combinations) that we chose. Nutrigenetics and nutrigenomics cover disparate fields and complex mechanisms that can be described in many different ways. Thus, our sample might not include all clinical studies that could have been identified with other key words and that could have met the three selection criteria mentioned above. Studies that only measured gene expression, without pointing out and referring to a particular polymorphism or specific DNA sequence(s) or variation were also excluded. We could only identify the participant exclusion or inclusion criteria that were explicitly reported by the authors in their publications or that could be inferred from population sample descriptions such as provided in the publications. Yet, only a full review of research protocols could give a real picture of the explicit exclusion or inclusion criteria used in nutrigenetics research. As mentioned above, 120 ‘referring’ publications referred those readers interested in getting more information about the participants or the methodology of the study to one or several previously published papers (‘referred publications’). Yet, 30 of these 120 referring publications provided us with one or more references, which we could not access (e.g. papers or books not accessible on the Internet or not accessible without fees from the electronic journals and library databases of our university). Moreover, 14 referring publications referred to an ongoing or previous study from which participants were recruited or in which participants took part but for which they did not provide any explicit reference in their bibliography. In 4 of these 120 referring publications, the references given for the description of the methodology, the participants and/or their selection were inaccurate (e.g. reference to the abstract of a poster at a symposium, an incomplete reference, a reference to a theoretical paper that contributed no


45

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Appendix I

useful information about the participants of the referring study, etc.). Finally, instead of providing the expected full description of the methodology of the study and/or of the participants, several referred papers turned to other previous publications for the same purpose (and in turn, some of the latter publications referred the readers to even earlier papers for the same descriptions). We did not extend our analysis of references beyond the references provided by referring publications. Finally, we cannot demonstrate that the articles in which we identified interpretative claims were actually read or used by the media – or the general public, or policymakers – and that they were therefore an actual source of biohype. Nevertheless, it is worth noting that these articles contributed to the development of the field of nutrigenetics in a manner similar to all the articles of our initial sample, if we refer to the number of times each article has been cited in the published literature and to the impact factor of the journals

References [1] National Institutes of Health, Office of Extramural Research: Glossary and Acronym List. 2011. http://grants.nih.gov/grants/glossary.htm. (Accessed September 2013). [2] Aronson, N. (1984). Science as claims-making activity: Implications for social problems research. In Studies in the Sociology of Social Problems. Schneider, J.W. and Kitsuse, J. I. (eds.). Norwood, N.J.: Ablex, pp. 1–30.

46


APPENDIX II QUESTIONNAIRE

Appendix II

Welcome to the OMICS-ETHICS research group online questionnaire. We are currently conducting a study to identify the scientific and ethical challenges in nutrigenomics/nutrigenetics (hereafter: "NGx") clinical research. As a basis for the study, the Omics-Ethics research group (University of Montreal, Department of social and preventive medicine) performed a systematic analysis of 173 studies published in this field. This analysis revealed various current or potential scientific and ethical challenges linked to NGx research and its current or future applications. However, an in-depth interpretation of the results of this analysis requires the expertise of researchers like you who are at the forefront of the development of NGx. All researchers invited to participate in this study are first authors or corresponding authors of NGx clinical studies. Your view is thus essential for this project and the Omics-Ethics research group invites you to fill in an online questionnaire, which should take you 30 minutes or less. Questionnaires will be anonymous and online data collection will be performed under strict security measures to avoid any breach of confidentiality. This research project is financed by the Fonds de recherche du Québec - Santé (FRQS) and the Canadian Institutes of Health Research (CIHR). All respondents will be sent a summary of the results and informed about any related publications. Your answers will be used for the purposes of this study only. By filling in the questionnaire, you consent to participate in this research project and agree with the use of the information you will provide such as described above. It will take you less than 30 minutes to fill in this questionnaire. You may pause the questionnaire at any time and resume later on. It is better not to use the "previous", "next" or "refresh" buttons of your web browser. Instead, please use the buttons included on each page of the questionnaire (e.g. : Next). For any information regarding the ethical conditions of your participation to this research project, or your participant's rights, or any complaint about your involvement in this study, or any other question on the project, please click here. 2Approval number from the Research Ethics Board and date of approval:

Project # 11T-020-CERES-D Approval date: December 22th, 2011 Any question on the project may be asked to Beatrice Godard, Ph.D., at the following address: Département de Médecine Sociale et Préventive, Université de Montréal, C.P. 6128, Succursale Centre-Ville, Montréal, QC, Canada, H3C 3J7 Phone : +1 514 343-6832 Fax : +1 514 343-2371 E-mail: [email protected] For any information regarding the ethical conditions of your participation to this research project, you may communicate with the Faculty of Medicine's Research ethics board's coordinator, at [email protected] or, by phone, at (514) 343-6111 ext. 2604. For more information about your participant's rights, you may consult the University of Montreal's webpage for participants : http://recherche.umontreal.ca/participants Any complaint about your involvement in this study can be addressed to the University of Montreal's ombudsman, at (514) 343-2100, or at the e-mail address [email protected] (accepts collect calls, speaks French and English, and receives calls from 9am to 5pm EST). We thank you for your participation! Click on the Access the survey button.

48

Appendix II Do you agree with the following definition of nutrigenomics? "Nutrigenomics studies the functional interactions of food and its components with the genome at the molecular and cellular level, as well as the systemic level, with the goal of using diet to prevent or treat diseases". (Curr Opin Lipidol. 15(2); 2004) Yes, totally Yes, partly No

Please provide comments, if any :

Do you agree with the following definition of nutrigenetics? "Nutrigenetics examines the effect of genetic variation on the interaction between diet and disease. Its goal is to generate recommendations regarding the risks and benefits of specific diets or dietary components to the individual". (Curr Opin Lipidol. 15(2); 2004)


4

Yes, totally Yes, partly No


0%

25%

50%

75%

100%

Please indicate your level of agreement on each of the following statements about nutrigenetics/nutrigenomics (hereafter "NGx")

I strongly agree

I generally agree

Still uncertain/cannot be answered at this stage

I somewhat disagree

I strongly disagree

NGx will provide health benefits to individuals who consume certain nutrients. NGx will turn food into medication. NGx will transform an enjoyable activity such as eating into a health hazard. NGx will allow the identification of individuals who suffer from adverse reactions caused by the consumption of certain nutrients. NGx will increase adherence to dietary advice thanks to individualized recommendations. NGx demonstrate that genetic risks and predispositions to complex diseases can be modified through dietary interventions. 5

NGx demonstrate that dietary interventions based on genotypic knowledge are a useful strategy for the prevention of some complex diseases. NGx will impede individuals' autonomy associated with daily food choices. Personalized dietary recommendations will become more effective through the incorporation of genetic information.

I strongly agree

I generally agree


I somewhat disagree

I strongly disagree

49

0%

25% 50%II 75% Appendix

100%

Please indicate your level of agreement on each of the following statements about nutrigenetics/nutrigenomics (hereafter "NGx")

I strongly agree

I generally agree


I somewhat disagree

I strongly disagree

Efficacy of preventive measures based on NGx will be uncertain because the long-term effects of changes in diet cannot be anticipated. Advances in the field of NGx enable more precise public health advice about dietary intake, supplement use, and genetic testing. NGx will place an excessive burden on individuals, because they will be responsible for being tested and remaining compliant with NGx dietary recommendations. More effective preventive and therapeutic interventions will be developed within the context of nutritional genomics as new tools become available to achieve effective dietary prevention and therapy. NGx will lead to the stigmatization of, or discrimination against, people who will not comply with NGx dietary recommendations.

7

Limitations resulting from methodological tools used in NGx clinical research (e.g.,questionnaires, study

design, etc.) raise concerns about the achievability of its benefits.

0%

25%

50%

75%

100%

The following statements describe potential benefits associated with the development of NGx. What is your opinion in regards to the plausibility of the achievement of each potential benefit? Already achieved

Likely achievable within 5 years

Likely not achievable within 5 years

Dietary interventions based on genotypic knowledge will be introduced as a useful strategy for the prevention of complex diseases. Personalized dietary recommendations will become more effective through the incorporation of genetic information. More precise public health advice about dietary intake, supplement use, and genetic testing will be possible through NGx research. NGx will allow the identification of individuals who suffer from adverse reactions caused by the consumption of certain nutrients. More effective preventive and therapeutic interventions will be developed within the context of nutritional genomics as new tools become available to achieve effective dietary prevention and therapy.

50


Likely not achievable

I do not know

0%

25%

50%

75%

Appendix II

100%

Please indicate your level of agreement with each of the following statements about ethnicity in NGx clinical research. I strongly agree

I generally agree

I somewhat disagree

I strongly disagree

A lack of representation of ethnic groups in NGx research limits the possibilities of generalizing the results to these populations. Ethnicity should be reported in scientific publications of NGx research, otherwise it cannot be inferred to what populations study findings can apply. The concepts of ethnicity, race or origin are vague and controversial. Thus, it is difficult to use them in NGx scientific publications.


| Interrupt

9

0%

25%

50%

75%

100%

What challenges, if any, do you see in the recruitment and the participation of ethnic minorities in NGx research? (You may choose several answers)

Adaptation of questionnaires to ethnic particularities Controversies about the use of "race", "ethnicity" or "ancestry" in biomedical research Measurement/assessment of dietary intake Measurement/assessment of ethnicity Compliance

Language Impact of population stratification on study results Obtaining informed consent Risks of stigmatisation or discrimination Cultural issues Other(s), please specify :

None

Willingness to participate Retention of participants

| Interrupt


51

0%

25%

50%

75%

Appendix II

100%

Please indicate your level of agreement on each of the following statements about participants' age in NGx clinical research. I strongly agree

I generally agree

I somewhat disagree

I strongly disagree

A lack of representation of infants, children and minors in NGx research limits the possibilities of generalizing the results to these populations. A lack of representation of the elderly in NGx research limits the possibilities of generalizing the results to these populations. If the age of participants is not reported in scientific publications of NGx research, it cannot be inferred to what populations study findings apply. There is no scientific rationale to include the elderly (75 years old and over) in NGx research.


11

0%

25%

50%

75%

100%

What challenges, if any, do you see in the recruitment and the participation of infants, children and minors in NGx research? (You may choose several answers)

Adaptation of questionnaires

Obtaining informed consent

Controversies about the notion of "minimal risk" or "acceptable burden" for children in clinical research

Ethical concerns about genetic testing on children

Assessment and measurement of dietary intake

Other(s), please specify :

Willingness to participate Compliance

Protection of children's autonomy Risks of stigmatisation or discrimination

None

Parental involvement Retention of participants

| Interrupt

52


12

Appendix II

0%

25%

50%

75%

100%

What challenges, if any, do you see in the recruitment and the participation of the elderly in NGx research? Comorbidities

Obtaining informed consent

Controversies about the notion of "minimal risk" or "acceptable burden" for vulnerable people in clinical research

Retention of participants

Pre-existing medication regimen

Adaptation of questionnaires

Willingness to participate

Other(s), please specify :

Measurement/assessment of dietary intake Involvement of relatives or surrogates

Protection of participants' autonomy Risks of stigmatisation or discrimination

None

Compliance

| Interrupt

13

0%

25%

50%

75%

100%

Have you ever encountered any ethical issues in conducting your NGx research projects? Yes No

Please briefly explain your answer :


53

14

Appendix II

0%

25%

50%

75%

100%

To what extent do you think it is important to include the following elements when reporting NGx study results in scientific publications : Very important

Somewhat important

Not very important

Not important at all

Eligibility criteria. A detailed description of the participants who were actually recruited (age, ethnicity, sex, comorbidities, etc.). Potential clinical applications that could result from the study findings. A discussion about the impact of methodological limitations on the results. A discussion about the impact of methodological limitations on the degree of achievability of the potential benefits of NGx.

| Interrupt 15

0%

25%

50%

75%

100%

What are your target audiences when you report NGx clinical research results in a scientific publication? You can provide the same answer for several audiences (e.g. two different audiences might be considered as first audiences.

First targeted audience(s)

Second targeted audience(s)

Third targeted audience(s)

General public. Health service providers (e.g., MDs, nutritionists, dieticians, etc.). Media. Peers/Colleagues. Policy makers.

| Interrupt

54


Fourth targeted audience(s)

Not a targeted audience

Appendix II

0%

25%

50%

75%

100%

Please indicate your level of agreement with each of the following statements about the reporting of potential benefits associated with the development of NGx in scientific publications (S examples of potential benefits : development of personalized dietary recommendations, more effective prevention strategies for complex diseases, more effective treatments, etc.). I strongly agree

I generally agree

I somewhat disagree

I strongly disagree

A manuscript has a better chance of being accepted for publication if the potential benefits associated with the development of NGx are stated. Statements about potential benefits (see some examples of these in question 13 above) in a scientific publication are useful to draw the readers' attention to the advantages of the development of NGx. Statements about potential benefits in a scientific publication foster continuing funding of NGx research. Statements about potential benefits in a scientific publication can improve the credibility of the publication. Benefits whose achievability is uncertain should not be mentioned in a scientific publication.


17

0%

25%

50%

75%

100%

For how long have you been doing research? years What is your area of expertise? (1 to 5 keywords)

How would you describe your current field of research? (1 to 5 keywords)


55

Appendix II

18

0%

25%

50%

75%

100%

In the last fifteen years, what were the sources of funding for the NGx research projects that you worked on (if any)? (You may choose more than one answer)

Public (government) funding agencies Universities Non-profit organizations (trusts, charities, foundations, etc.) Private companies Other(s), please specify :

Not applicable

| Interrupt 19

0%

25%

50%

75%

100%

In what country are you based?

Where do you conduct your NGx researches?

| Interrupt

56


Appendix II

20

0%

25%

50%

75%

100%

75%

100%

What kind of institutions are you affiliated with? (You may choose more than one answer)

Academic (University) Public hospital Private hospital/clinic Private laboratory or company Non-profit organization (research organisation, trust, charity, etc.) Other(s), please specify :

What is your current status? (You may choose more than one answer)

PhD student Postdoctoral student Assistant professor/researcher

21

Associate professor/researcher Full professor M.D. Other(s), please specify :

0%

25%

50%

What is your gender? Female Male What age group do you belong to? 30 and under 31-40 41-50 51-60 61 and over

| Interrupt


57

APPENDIX III RESPONDENTS’ COMMENTS ON THE DEFINITION OF NUTRIGENOMICS

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions APPENDIX III RESPONDENTS’ COMMENTS ON THE DEFINITION OF NUTRIGENOMICS

In all, 42 respondents out of 126 (i.e. 33%) provided comments to explain their answer to the following question: Do you agree with the following definition of nutrigenomics? ‘‘Nutrigenomics studies the functional interactions of food and its components with the genome at the molecular and cellular level, as well as the systemic level, with the goal of using diet to prevent or treat diseases.’’ (Curr Opin Lipidol. 15(2); 2004) *Legend: 1: Yes totally 2: Yes, partly 3: No All comments describing doubts or concerns about the goal of nutrigenetics as outlined in the definition, or proposing a definition that does not include a goal of this sort, can be found in yellow cells.

60



Comments

2

Add a goal to optimize mental and physical performance.

3

It’s not just the interaction of food with the genome, but how the genome influences the way food affects body functions.

2

Also to understand how it works!

2

Manifestation of a disease is partly predisposed by the inter-molecular dependence in the natural genetic makeup, which can predispose to a disease, say up to 40%. However 60% of the outcome can be influenced, promoted or protected by influence of the nutrition and environment. Disease shall manifest easily in an individual with genetic pre-disposition. Can be prevented by diet & lifestyle modification.

2

It does not only study the functional interactions but all the interactions, functional or not. Tools exist to study the whole underlying mechanism either if this could help to prevent disease or not.

2

The goal may be broader than preventing or treating diseases.

3

Far too restrictive. This also includes nutrigenetics. The goal I would see is to understand homeostasis and the beginnings of diseases, far more than treat them.

2

I would include some reference to epigenomics.

2

Nutrigenomics is the biological response in response to food components as influenced by genetic, epigenetics and transcriptomics.

2

The more fruitful approach would be to focus on molecular nutrition because it is unlogical to separate out just the interactions between nutrients/food components with genes. Nutrient-gene interaction is not enough. We need understanding of how nutrients work with & without genes. Mechanisms of action are required. Can apply principles in new situations.

2

The importance of high-throughput technologies is central to Nutrigenomics, as opposed to simple nutritional sciences. The goal also should be to reduce the risk of disease.

2

1) It is not only to treat disease, but especially to improve health. 2) Interaction with the genome suggests direct interactions only. In my opinion it is about the expression of the genome (at different levels; mRNA, protein, metabolite, etc.)

2

I think that nutrigenomics is not only applied but also basic research. Therefore the second part of this definition is not obligatory.

2

I will add to maintain health, to understand physiology. There are several aims in nutrigenomics.

2

It does not always need to be focused on disease prevention or treatment.

2

I don’t assume ‘‘prevent or treat disease’’ is part of the definition.

2

I don’t understand the ‘‘systemic level’’, the rest is OK.


Appendix III

*

61

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Appendix III 62

*

Comments

2

Should not be focused only to disease.

2

This definition excludes understand the relationship between food and our genome without particular reference to disease or achieving optimal health.

1

The above definition seems very appropriate as it includes gene-diet interactions in the prevention and possibly management of common disorders such as type 2 diabetes mellitus, cardiovascular disease, cancers etc. Due to the necessity of daily consumption, food constitutes the most important environmental factor in diseases triggered by gene-environment interactions. Whereas understanding the impact of dietary micro- or macronutrients upon gene expression may help in the prevention or treatment of common diseases whose underlying aetiology is gene-diet interaction, the effect of dietary components in people with variant forms of gene(s) may pave the way for personalized diet to aid in the prevention of common disorders.

2

With the goal of using diet to prevent or treat diseases, and to benefit health both individually or “populationally”.

2

I would omit the goal.

2

In my opinion Nutrigenomics studies biomarkers of interactions of the diet and the genome. I do not see any evidence that the research has advanced to determine the functional consequences. However, I do hope it will happen in the immediate future.

2

For me this concept should be expanded on how diet affects an individual’s genes and how genetic makeup can affect the way a person responds to diet. Not only for the curing of diseases, but also, for example, to increase athletic performance and achieve a better quality of life.

2

The goal of nutrigenomics is to understand the interactions.

2

This is one of the goal but other goals exist (fundamental knowledge, other effects of food and its components...).

2

The genome also acts, through proteins and RNA expression, on the metabolism of nutrients. It is not a one-way interaction.

2

It is not necessarily to treat or prevent anything - it is also for understanding how this interplay works.

2

The definition is broader - it is ‘‘to understand the fundamental interactions of food and its components with the genome at the molecular and cellular level, as well as the systemic level, and to enable diet to be used to prevent or treat diseases’’. In other words it is for fundamental understanding as well as with a practical goal.

1

Nutrigenomics studies the effects of nutrients on gene expression - See Simopoulos’ definition.

2

Also to understand the systems response to diet, not only to prevent or treat diseases.

3

A very dated definition... most of my colleagues use the term to reflect nutrient-gene interaction with a focus on transcriptomics and with a health rather than disease focus.



Comments

3

The definition of Nutrigenomics as developed by the International Society is the role of diet or nutrients on gene expression.

2

Goal can be understanding (though ideally knowledge generated is applicable).

2

I would actually tend to reverse the definitions of Nutrigenomics and Nutrigenetics as outlined here, to some extent. Nutrigenomics to me is more about the interaction between diet and the genome, and its effect on disease, while Nutrigenetics focuses more closely on specific functional interactions.

2

With the goal.... should be omitted, because nutrigenomics could serve other purposes as well.

2

There is not necessarily the disease focus of the goal.

2

[Comment removed as it may identify the respondent and cannot be paraphrased.]

3

High-throughput analysis of the genome is missing.

2

I agree if ‘systemic level’ means that effects downstream of the genome are included.

2

Health is more than an absence of disease. And nutrigenomics could help to find a healthier (and happier) personal life based on differences in the genome and how it is regulated and expressed.

2

A more concise definition would be better e.g. ‘‘Nutrigenomics studies the influence of nutrition on gene expression’’.


Appendix III

*

63

APPENDIX IV RESPONDENTS’ COMMENTS ON THE DEFINITION OF NUTRIGENETICS

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions APPENDIX IV RESPONDENTS’ COMMENTS ON THE DEFINITION OF NUTRIGENETICS

In all, 43 respondents out of 126 (i.e. 34%) provided comments to explain their answer to the following question: Do you agree with the following definition of nutrigenetics? ‘‘Nutrigenetics examines the effect of genetic variation on the interaction between diet and disease. Its goal is to generate recommendations regarding the risks and benefits of specific diets or dietary components to the individual.’’ (Curr Opin Lipidol. 15(2); 2004) *Legend: 1: Yes totally 2: Yes, partly 3: No All comments describing doubts or concerns about the goal of nutrigenetics as outlined in the definition, or proposing a definition that does not include a goal of this sort, can be found in yellow cells.

66



Comments

2

Again, I would encourage consideration of metrics different from disease entities.

3

It’s not just about disease, its about physiology in a broader sense, how food affects the functioning of the body.

2

Also to understand mechanisms, because the aim of specific recommendations will not be possible to reach for many diseases in which the genetic basis bears on multiple genetic variants with very small effects (even in combinations, these effects can only be seen at the population level but not at the individual level).

2

I consider this the interactions of genes and nutrition in the occurrence of disease.

2

I believe this definition is too narrow. First, it ignores the contribution of behavioural genetics. Genetics influences food choices and health beyond metabolic aspects. Second, the field need not be limited to personalized recommendations to have utility.

2

I don’t believe that the definition should be restricted to interactions between diet and disease. In my opinion, nutrigenetics can also consist of studying the effect of genetic variation on the metabolism of foods.

2

I would replace ‘‘disease’’ with a broader term such as ‘‘phenotypes’’.

2

Not very up to date definitions.

2

Nutrigenetics examines the effect of genetic variation on the interaction between diet and ‘‘disease’’. I think it is better to use health-related events rather than disease.

2

There needs to be a statement about heritability and/or germline DNA so that it is clear that ‘genetics’ implies DNA that is present in all cells of the body.

2

I do not entirely agree with the stated goal. Rather I would say the objective is to identify potential risk and benefits of diet or dietary components for a subset of individuals with respect to the disease of interest.

2

It is not diet per se but bioactive food components and it is likely health promotion and disease prevention.

1

Turns out to be very complicated due to the presence of millions of SNPs.

2

The goal is also to obtain understanding.

2

As above [i.e. same comment as about the definition of nutrigenomics: “I think that nutrigenomics is not only applied but also basic research. Therefore the second part of this definition is not obligatory”.]

2

Nutrigenetics is the study how the genome affects responses to diet and also how the diet could affect the genome. Goal is to generate recommendations, to stratify cohorts to better understand mechanisms of actions, etc.

3

Goal to assess impact of genetics on nutrition not to provide individuals with advice.

2

Should not be focused only on disease.

2

This definition excludes understand the relationship between our genetics and diet and lifestyle choices, again without particular reference to disease or achieving optimal health. © Omics-Ethics Research Group 2013

Appendix IV

*

67

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Appendix IV 68

*

Comments

1

This is again true. The concept of ‘nutrigenetics’ and its application would enable formulation of specific diets for an individual and personalized counselling with the aim of reducing the risk of a given disorder.

2

I would omit the goal.

1

Why not keep it simple: The study of variations of DNA and RNA characteristics as related to nutrient responses.

2

Modified definition: ‘‘Nutrigenetics examines the impact of genetic variation on the association between diet and disease. Its goal is to generate recommendations regarding the risks and benefits of specific diets or dietary components to the individual’’.

2

See comments above. [i.e. same comment as about the definition of nutrigenomics: “For me this concept should be expanded on how diet affects an individual’s genes and how genetic makeup can affect the way a person responds to diet. Not only for the curing of diseases, but also, for example, to increase athletic performance and achieve a better quality of life”.]

2

The goal of nutrigenetics is to understand the interactions.

2

Primary goal would be to generate understanding/scientific knowledge.

2

The focus of nutrigenetics should be on the study of individual genes, while nutrigenomics studies all genes.

2

Similar response [i.e. same comment as about the definition of nutrigenomics] I would say ‘‘its main goal’’.

3

There is no such thing as nutrigenetics. All interactions between the environment and the genome and the genome and the environment involve many aspects of the genome: single nucleotide variants, copy number variants, DNA methylation, epistatic (gene gene) interactions. Nutrigenetics implies we can identify single SNPs in genes that can be used to deterministically predict response. The response will be probabilistic.

2

See above [i.e. same comment as about the definition of nutrigenomics: it is not necessarily to treat or prevent anything - it is also for understanding how this interplay works.]

3

Same comment and principle as above [i.e. same comment as about the definition of nutrigenomics: The definition is broader - it is ‘‘to understand the fundamental interactions of food and its components with the genome at the molecular and cellular level, as well as the systemic level, and to enable diet to be used to prevent or treat diseases’’. In other words it is for fundamental understanding as well as with a practical goal.] - this definition is very skewed to one goal only.

2

Nutrigenetics examines the effect of genetic variation on the interaction between diet and phenotype (not only disease).

2

A bit limiting, SNPs are but one form of variation, copy number could be just as important.

3

Nutrigenetics is the study of individuals and their genetic variants to dietary response.



Comments

2

Recommendations can apply to groups of individuals not necessarily to a given individual.

3

Modification of the relation between diet and disease by genetic variation, or of the relation between gene and disease by diet. -subgroups (not individual).

2

As above [i.e. same comment as about the definition of nutrigenomics: “I would actually tend to reverse the definitions of Nutrigenomics and Nutrigenetics as outlined here, to some extent. Nutrigenomics to me is more about the interaction between diet and the genome, and its effect on disease, while Nutrigenetics focuses more closely on specific functional interactions”.]

2

Same as above [i.e. same comment as about the definition of nutrigenomics:] ‘‘its goal...’’ is not relevant for the definition, because it is just one of the possible goals and may reflect a personal preference of the author(s)

2

It is rather the interaction between genetic variants and diet in determining risk of disease or intermediate steps on the way to disease.

2

I do not agree with the goal to generate recommendations: far too premature and probably unreachable given the complexity of gene-gene and genetic-epigenetic interactions and the lack of clinical studies to assess all these interactions to end up with recommendations at the level of the individual.

2

Same thing primarily. Health is more than absence of disease. Also the nutrigenetics and nutrigenomics are highly interrelated, thus you need one to be able to do the other to a qualitatively high extend. Meaning they also share goals.

3

I’m missing the ‘‘system biology’’ in this definition.

2

A more concise definition would be better e.g. ‘‘Nutrigenetics studies the influence of genetic variability on responses to nutrition’’.


Appendix IV

*

69

APPENDIX V RESPONDENTS’ COMMENTS ON THE STATEMENTS ABOUT ETHNICITY OF PARTICIPANTS IN NGX RESEARCH

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions APPENDIX V RESPONDENTS’ COMMENTS ON THE STATEMENTS ABOUT ETHNICITY OF PARTICIPANTS IN NGx RESEARCH In all, 22 respondents out of 126 (i.e. 17%) provided comments to explain their answer to the following questions about the ethnicity of participants in NGx research:

Please indicate your level of agreement with each of the following statements about ethnicity in NGx clinical research: Statement A: “A lack of representation of ethnic groups in NGx research limits the possibilities of generalizing the results to these populations.” Statement B: “Ethnicity should be reported in scientific publications of NGx research, otherwise it cannot be inferred to what populations study findings can apply.” Statement C: “The concepts of ethnicity, race or origin are vague and controversial. Thus, it is difficult to use them in NGx scientific publications.” Strongly or somewhat disagree Strongly or generally agree

72



B

C

Comments The concept of ethnicity or ethnic group is controversial and can be negatively perceived from a general point of view. However, if we restrict the concept of ethnicity to the field of genetics, projects like HapMap and more recently the 1000 Genome project have demonstrated the reality of distinct ethnic groups at the genome level.

Appendix V

A

[… we suggest to use] ‘‘ethnogeographic origin’’ because it may be a more accurate descriptor for genomic ancestry. The concepts of ethnicity etc. are vague and controversial, nevertheless they are confounding factors in genetic studies of all kinds! A lot of useful guidelines can be achieved in spite of the present limitations. One aspect that is poorly discussed is the idea of 1st generation, 2nd generation, etc immigrants. It is clear that the environment influences disease risk, and that these risks differ between ethnicities. However, are 1st generation immigrants more susceptible to these risks than subsequent generations? This is unknown and could serve as a significant confounder in nutrigenomics research. Until further knowledge and/or discoveries, the ethnicity should be mentioned in every trial involving humans. Must consider population stratification that can occur in multiethnic population when conducting gene-environment research. Best replicated finding within several different populations. Context matters. I agree that the concepts of ethnicity, race or origin are vague and controversial. It is still feasible to use them in NGx scientific publications; however, the reader needs to interpret the results with caution (and not all readers will do this). However, if NGx data is to be of use in a public health capacity, this issue will have to be overcome. The ‘‘embarrassment’’ of racial stereotyping does, in my opinion, cause problems in interpreting data. There are unquestionably differences in genetic make up and metabolism between different races, and not taking this into account in NGx is fatuous Ethnicity is not different from genotype. Ethnicity should be reported for better conducting nutrigenetics studies, but anonymity has to be ensured like with all other clinical data. Food preferences/availability and metabolism reflect geographical and climatic area and thus inevitably involve ethnicity - disregarding this can lead to false recommendations.


73

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Appendix V

A

B

C

Comments Food habits are usually specific for different ethnic/regional populations. Hence, although the concept of ethnicity may be debatable from a genetics point of view, this may not be so while addressing gene-diet interactions. Thus it should be relevant to report ethnicity of the study group in NGx scientific publications. Many aspects of genetics should be considered and analysed when studying diet - gene interactions such as epigenetics and transcriptomes ideally in a long-term study. There are clearly differences in the distribution of inherited factors between population groups, e.g SNP distribution and LD [sic] between Caucasian and African. This leads to differential responses, which needs to be determined for each population group. However, one always needs to be aware that this by no means makes on group inferior/superior. Ethnicity is known to affect BMR, it is very likely to affect NGx. More information about the study populations is always better. Nutrigenomics research, indeed all biomedical research, must be redesigned to take into account differences in genetic makeup and environmental exposure differences. It has little to do with ethnicity, and more to do with individual genetic variation. There are accepted terms for ethnicity/ race/ etc. based on Census data. Although it is somewhat of a grey area, in terms of validity, it may be quite relevant to report ethnicity. This needs to be weighed against its controversiality, which may also be addressed and lessened by transparency and better communication to non-scientists. This depends on the country. Some populations are quite homogenous, others are quite mixed, so whenever it gives sense, it should be reported.

74


APPENDIX VI RESPONDENTS’ COMMENTS ON THE STATEMENTS ABOUT THE AGE OF PARTICIPANTS IN NGx RESEARCH

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions APPENDIX VI RESPONDENTS’ COMMENTS ON THE STATEMENTS ABOUT THE AGE OF PARTICIPANTS IN NGx RESEARCH In all, 15 respondents out of 126 (i.e. 12%) provided comments to explain their answer to the following questions about the age of participants in NGx research:

Please indicate your level of agreement with each of the following statements about ethnicity in NGx clinical research: Statement D: “A lack of representation of infants, children and minors in NGx research limits the possibilities of generalizing the results to these populations.” Statement E: “A lack of representation of the elderly in NGx research limits the possibilities of generalizing the results to these populations.” Statement F: “There is no scientific rationale to include the elderly (75 years old and over) in NGx research.” Statement G: “If the age of participants is not reported in scientific publications of NGx research, it cannot be inferred to what populations study findings apply.”

Strongly or somewhat disagree Strongly or generally agree

76



E

F

G

Comments All age groups should initially be included, later a set of guidelines can be applicable to a population. I suppose the underlying thought to not include the elderly in NGx research is because it would be ‘‘too late’’ to benefit from lifestyle changes; however, there is no indication that benefits can’t be experienced in the short term. There is a complete lack of research exploring whether genetic information can instil lifestyle changes - thus the speed at which improvements in health are experienced remains unknown.

Appendix VI

D

Age is important to consider in genetic research - can have survivor bias The limitations are pretty much the same with and without genetic stratification. We are in danger of overemphasizing the problems and lose sight of the same uncertainties with any question. Why should we be more sceptical about estimating genotype-specific nutrient requirements but be comfortable with non-stratified outcomes? We now know about enough instances where genotypic variation in groups covers up very real nutrient effects in defined subgroups. The results should not be generalized. The impact of gene polymorphism on nutritional phenotype can change over time. As with dietary recommendation it is important to include subjects with a wide range of ages, as for certain physiological aspects age and gender are the main factor of individual variability. Yet, work with infants, children or minors has to be as reduced as possible. Including the elderly - it depends which kind of results one is looking for (immediate effects vs. long-term) Humans who achieve great age are a biologically-fascinating. They offer an opportunity to examine the opposing scenario to the much examined early onset diet-related ill-health. What has enabled these individuals to achieve great age is as important if not more so than why others do not. It is important comparisons between young and older adults, since most complex diseases manifest later. Age is important, but the results cannot be generalized to infants, children, etc, since it is not only about age as a classifier, but also the individual at the age. Nevertheless, our new experimental design can take into account age and individual assessments. The study design based on pre-selected groups (based on age, ethnicity, or any other apparent phenotype) are highly flawed given that each individual is genetically different and has a unique environment.


77

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Appendix VI

D

E

F

G

Comments If the assertions behind the questions here are correct - I do not know if they are. What has age got to do with genetic testing...assuming you mean a SNP rather than epigenetic change. This is a generic question...lack specificity to GTX A lack of representation of young people also limits the possibilities of studying long term/lifetime effects. As for many other questions in this survey, these issues are not specific to NGx but general for nutritional research, which is making the questions difficult to answer as they are posed. 75 years is in fact no longer old. These people might have over 25 years to live and healthy diet might help from them. Apart from that they might provide interesting mechanistic clues

78


APPENDIX VII RESPONDENTS’ COMMENTS ON THE ETHICAL ISSUES ENCOUNTERED IN NGx RESEARCH

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions APPENDIX VII RESPONDENTS’ COMMENTS ON ETHICAL ISSUES ENCOUNTERED IN NGx RESEARCH

In all, 29 respondents out of 126 (23%) answered that they had already encountered some ethical issues while conducting their NGx projects. Among them, two respondents did not comment on or explain their answer. Respondents’ comments can be classified into 5 categories (see Table 1). All comments can be found in Table 2.

Table 1 Category number 1. 2. 3. 4. 5.

Ethical issues Issues linked to the disclosure of individual incidental and/or research results (including disclosure to family members) and more generally to the identification of genetically at risk participants in NGx research Issues linked to informed consent Difficulties to get ethical approval or issues raised by Research Ethics Boards Issues linked to ethnicity or genetic variations across populations Others

Number of comments 13 5 5 3 4

Table 1: Researchers’ comments on ethical issues they encountered while conducting NGx research. Twenty-seven respondents commented their answer (out of 29 respondents who stated having encountered some ethical issues). When appropriate, certain comments were split into specific parts that could be classified into different categories. The same comment could also appear in more than one category. Thus the total number of comments in this Table is > 27.

80



1. Issues linked to the disclosure of individual incidental and/or research results (including disclosure to family members) and more generally to the identification of genetically at risk participants in NGx research

Appendix VII

Table 2

Discussion with scientific journal whether we had to report back individual results of SNP-testing while participants consented to not being informed about individual results. Question of informing experimental subjects with outlying data on biomarkers when no real concept of what this may mean in terms of health or disease risk. Biological patterns that differ significantly from other participants, which may indicate as yet undiagnosed disease, which enable researchers to ‘identify’ the individual’s samples albeit not the individual in the study - issues arising: (1) should the individual be identified (i.e. the study code broken) and referred to the family doctor, (2) a double-blinded study is not valid if researchers are able to identify specific samples. We identified in proteomics profiling a volunteer with a rare disease to which currently very little information is available. This incidental finding caused a lot of friction with ethics. Problem to define the communicable preliminary results. We often deal with the question of informing participants of genotypes that we feel may be pertinent to their health, for example we studied ABCG 8 mutations which can increase ‘‘absorption’’ of dietary plant sterols which are thought to potentially be detrimental to health. Disclosure issues to family members not participating in the study. Difficulty of explaining to participants that we would measure a characteristic of their genetic material but would only report back results on group level. We did not think results on an individual level would be meaningful and we did not know whether participants would be able to deal with their individual results and so made this decision for them. ‘Omics research is generally explorative, so it can be difficult to ask volunteers to participate in projects where you are searching for markers related to risk, yet you cannot tell the interested participant afterwards whether she/he is at any in- or decreased risk. This relates to research on apoE genotyping long before the strong link with late-onset AD was known and even the relevance with CVD was in doubt. The later discoveries of the (multiple) disease links drove home the point that we can never assume that any genetic information is benign. Now, more than 30 years later I know better. Risk of causing harm to sick people/patients. Risk of causing harm to genetically at risk but overtly healthy individuals. […] Metabolic phenotyping of participants in a nutrigenomics weight loss intervention revealed several outliers of analytes indicating (risk for) disease. We found clear outliers in general studies, which seemed to indicate disease. © Omics-Ethics Research Group 2013

81

Challenges in nutrigenomics / nutrigenetics – Researchers’ perceptions Appendix VII

Table 2 (cont’d) 2. Issues linked to informed consent The consent for using human tissues. Discussion with scientific journal whether we had to report back individual results of SNP-testing while participants consented to not being informed about individual results. Informed consent use high validity data. Explaining and obtaining informed consent when the outcome is uncertain. In coordination of a European cross-national trial, it became obvious that there are quite different ethical rules in the various countries about what is ethically acceptable and how a certain informed consent shall be interpreted. 3. Difficulties to get ethical approval or issues raised by Research Ethics Boards I wanted to recruit Caucasians only (reduce genetic and diet variability) but this issue was raised by the Ethics board. - How do we assess Caucasians? - They were not convinced that we should exclude other ethnic groups. The research project was reviewed by the IRB. [The respondent indicated having faced ethical issues; thus, we inferred from his/her comment that such issues were raised by the Research Ethics Board.] Some Research Ethics Boards ask which specific genetic markers will be examined and how long will samples be kept. Restricting markers to ones that can be identified now and limiting the time that samples are stored impedes progress. […] the ethical issues are red herrings brought up by institutional review boards who assume that they know what is best for individuals. Most nutrigenomic studies are of very low risk by definition: we are testing generally regarded as safe foods. Yet the human study boards feel they have to ‘‘protect’’ the subjects... Informed consent means that the individual understands the risks and accepts them. The IRB’s often interject their own fears into the process. In coordination of a European cross-national trial, it became obvious that there are quite different ethical rules in the various countries about what is ethically acceptable and how a certain informed consent shall be interpreted. 4. Issues linked to ethnicity or genetic variations across populations Backgrounds of populations are different. What we have learned from the studies of the adiponectin gene, the genetic polymorphisms are found to be associated with type 2 diabetes and obesity in many populations. But, the associated polymorphisms are different in different populations. [Comment removed as it may identify the respondent and cannot be paraphrased.] I wanted to recruit Caucasians only (reduce genetic and diet variability) but this issue was raised by the Ethics board. - How do we assess Caucasians? - They were not convinced that we should exclude other ethnic groups.

82


Table 2 (cont’d) 5. Others Population scepticism against everything associated with genes. Compliance and assessment of foods intake. […] Causing unjustified or unnecessary suffering to animal disease models. Any animal and human study has ethical issues.

83