Comparison of Nutrigenomics Technology Interface Tools for Consumers and Health Professionals: A Sequential Explanatory Mixed Methods Investigation

Background Nutrigenomics forms the basis of personalized nutrition by customizing an individual’s dietary plan based on the integration of life stage, current health status, and genome information. Some common genes that are included in nutrition-based multigene test panels include CYP1A2 (rate of caffeine break down), MTHFR (folate usage), NOS3 (risk of elevated triglyceride levels related to omega-3 fat intake), and ACE (blood pressure response in related to sodium intake). The complexity of gene test–based personalized nutrition presents barriers to its implementation. Objective This study aimed to compare a self-driven approach to gene test–based nutrition education versus an integrated practitioner-facilitated method to help develop improved interface tools for personalized nutrition practice. Methods A sequential, explanatory mixed methods investigation of 55 healthy adults (35 to 55 years) was conducted that included (1) a 9-week randomized controlled trial where participants were randomized to receive a standard nutrition-based gene test report (control; n=19) or a practitioner-facilitated personalized nutrition intervention (intervention; n=36) and (2) an interpretative thematic analysis of focus group interview data. Outcome measures included differences in the diet quality score (Healthy Eating Index–Canadian [HEI-C]; proportion [%] of calories from total fat, saturated fat, and sugar; omega 3 fatty acid intake [grams]; sodium intake [milligrams]); as well as health-related quality of life (HRQoL) scale score. Results Of the 55 (55/58 enrolled, 95%) participants who completed the study, most were aged between 40 and 51 years (n=37, 67%), were female (n=41, 75%), and earned a high household income (n=32, 58%). Compared with baseline measures, group differences were found for the percentage of calories from total fat (mean difference [MD]=−5.1%; Wilks lambda (λ)=0.817, F1,53=11.68; P=.001; eta-squared [η²]=0.183) and saturated fat (MD=−1.7%; λ=0.816; F1,53=11.71; P=.001; η²=0.18) as well as HRQoL scores (MD=8.1 points; λ=0.914; F1,53=4.92; P=.03; η²=0.086) compared with week 9 postintervention measures. Interactions of time-by-group assignment were found for sodium intakes (λ=0.846; F1,53=9.47; P=.003; η²=0.15) and HEI-C scores (λ=0.660; F1,53=27.43; P<.001; η²=0.35). An analysis of phenotypic and genotypic information by group assignment found improved total fat (MD=−5%; λ=0.815; F1,51=11.36; P=.001; η²=0.19) and saturated fat (MD=−1.3%; λ=0.822; F1,51=10.86; P=.002; η²=0.18) intakes. Time-by-group interactions were found for sodium (λ=0.844; F3,51=3.09; P=.04; η²=0.16); a post hoc analysis showed pre/post differences for those in the intervention group that did (preintervention mean 3611 mg, 95% CI 3039-4182; postintervention mean 2135 mg, 95% CI 1564-2705) and did not have the gene risk variant (preintervention mean 3722 mg, 95% CI 2949-4496; postintervention mean 2071 mg, 95% CI 1299-2843). Pre- and postdifferences related to the Dietary Reference Intakes showed increases in the proportion of intervention participants within the acceptable macronutrient distribution ranges for fat (pre/post mean difference=41.2%; P=.02). Analysis of textual data revealed 3 categories of feedback: (1) translation of nutrition-related gene test information to action; (2) facilitation of eating behavior change, particularly for the macronutrients and sodium; and (3) directives for future personalized nutrition practice. Conclusions Although improvements were observed in both groups, healthy adults appear to derive more health benefits from practitioner-led personalized nutrition interventions. Further work is needed to better facilitate positive changes in micronutrient intakes. Trial Registration ClinicalTrials.gov NCT03310814; http://clinicaltrials.gov/ct2/show/NCT03310814 International Registered Report Identifier (IRRID) RR2-10.2196/resprot.9846

Although the advancement of nutrigenomics-based personalized nutrition shows signifcant promise in improving population health, it also presents challenges. These issues include concerns about the complexity in translating gene-based results into meaningful recommendations that will lead to positive health outcomes [13][14][15]. Differences in dietary intakes are not always observed between ``risk`` and ``nonrisk`` groups [20] and dietary changes have not been consistently observed across all identified risk gene variants (e.g., MTHFR) where nutrition advice is provided [16][17][18][19]. In order for nutrigenomics and personalized nutrition to advance in health practice, better interface educational tools (e.g., web applications, targeted messaging after personalized nutrition advice provided) need to be developed that are easily implemented by practitioners, understood by consumers, and that foster positive eating behavior changes. Furthermore, they need to incorporate accepted nutrition guidelines, integrate phenotypic information about current health status, and align with behavior change theory principles [5,20]. 2a-ii) Scientific background, rationale: What is known about the (type of) system Since the success of the Human Genome project, science technology has advanced rapidly in several disciplines, including medicine and nutrition. Given that the interaction of nutrients with DNA can impact nutritional status and the development of complex diseases, nutritional genomics (nutrigenomics) has become an increasingly important in nutrition practice. Nutrigenomics encompasses nutrigenetics, which investigates the effect of genetic variation on nutrient bioavailability and metabolism, and nutrigenomics, which examines how nutrients and bioactive food compounds affect human health through epigenetic modifications [1][2][3][4][5].

Does your paper address CONSORT subitem 2b?
The main study objectives were to compare a practitioner-facilitated personalized dietary approach that uses genotypic and phenotypic information to a self-driven approach and their impacts on changing participant's knowledge, motivation, and behavior related to eating habits and the quality of their diet. It was hypothesized that signifcantly higher levels of knowledge, motivation, and behavior would be reported and that there would be a higher level of diet quality changes in the group that receives personal DNA diet information plus customized dietary advice (practitioner led) compared to the group that is provided personal DNA diet information (direct-to-consumer self-driven approach) only. In addition, self-effcacy and quality of life measures were evaluated as potential mediators/moderators of dietary changes and outcomes.

3a) CONSORT: Description of trial design (such as parallel, factorial) including allocation ratio
The sequential explanatory mixed methods investigation consisted of: 1) a randomized controlled trial (RCT: 2:1 allocation ratio) comparing standard selfdriven versus practitioner-facilitated approaches that use DNA-based diet information; and 2) qualitative investigation of participants' experiences to help interpret the intervention's quantitative outcomes. 3b) CONSORT: Important changes to methods after trial commencement (such as eligibility criteria), with reasons The study protocol, including paper-based on online data collection forms, was approved by Quorum Institutional Review Board (protocol # 32220CDN/1). No changes were made 3b-i) Bug fixes, Downtimes, Content Changes Not applicable 4a) CONSORT: Eligibility criteria for participants Yes -information provided in multimedia appendex 4a-i) Computer / Internet literacy Eligibility criteria included that they wanted to improve their health, could understand and provide informed consent, and were willing to provide a buccal swab for DNA testing. Exclusion criteria are specified in (Multimedia appendix 1. Supplementary file 1. 4a-ii) Open vs. closed, web-based vs. face-to-face assessments: Participants selected for the study included healthy medically stable adults (35-55 years) residing in the greater Vancouver area of the province of British Columbia, Canada. Participants were recruited via social media, a newspaper article, and posters.

4a-iii) Information giving during recruitment
All participants provided initial online consent to collect eligibility screening information and, if eligible, baseline information related to their health. At the participant's first site visit, a second written consent form was reviewed which participants signed to confirm their continued involvement in the study. 4b) CONSORT: Settings and locations where the data were collected Participants selected for the study included healthy medically stable adults (35-55 years) residing in the greater Vancouver area of the province of British Columbia, Canada. Participants were recruited via social media, a newspaper article, and posters. 4b-i) Report if outcomes were (self-)assessed through online questionnaires Recruiting/Screening: This included the initial eligibility screen and, where applicable, baseline assessment (online) that collected information about sociodemographics, current health status (e.g., presence of health conditions, medication and supplement usage), quality of life, self-e cacy, questions about knowledge, motivation, and action related to DNA-based information, stage of change, physical and sedentary activities, food intakes (food frequency, food selection), anthropometrics, and sleep quality. 4. Follow-up #1. An online survey was sent to I and C participants at week 3 postintervention to collect data about any changes in income, social support, changes in knowledge, behavior, and action, stage of change, and adverse event information. For the I group, questions that asked about whether knowing one's personal DNA helped with eating behavior change were also included. 5. Follow-up #2. Six weeks after the intervention participants received the online baseline health assessment questionnaire and food records to complete in preparation for the final on-site visit (week 8 post-intervention). 4b-ii) Report how institutional affiliations are displayed Not applicable 5) CONSORT: Describe the interventions for each group with sufficient details to allow replication, including how and when they were actually administered 5-i) Mention names, credential, affiliations of the developers, sponsors, and owners Not applicable 5-ii) Describe the history/development process Not applicable 5-iii) Revisions and updating Not applicable 5-iv) Quality assurance methods Quantitative Analysis Food Intake and Nutrient Analysis: Nutrient analysis was conducted using ESHA -The Food Processor Nutrition Analysis and Fitness software and the Canadian Nutrient File [41,42]. Averages of the three days of nutrient values were used in the analysis. FFQ values were used to derive usual intakes of the nutrients of interest [24]. All analyses were done on an intent-to-treat basis using STATA software [43]. Qualitative Analysis Textual data from the online questionnaires (e.g., participant's personal dietary goals) were grouped into categories where feasible. Data from the focus groups were transcribed, entered into NVivo [44], and analyzed by research team members using interpretative thematic analysis to identify patterns, concepts, themes, and examples in relation to existing behavior change theories and the study objectives [45]. Interpretations were reviewed by research team members and participants to check for descriptive and interpretative validity. Qualitative data was reported based on thematic analysis derived from three independent reviews of the textual data.

5-v) Ensure replicability by publishing the source code, and/or providing screenshots/screen-capture video, and/or providing flowcharts of the algorithms used
Not applicable 5-vi) Digital preservation Not applicable 5-vii) Access Not applicable 5-viii) Mode of delivery, features/functionalities/components of the intervention and comparator, and the theoretical framework Not applicable 5-ix) Describe use parameters Not applicable

5-x) Clarify the level of human involvement
The sequential explanatory mixed methods investigation consisted of: 1) a randomized controlled trial (RCT: 2:1 allocation ratio) comparing standard selfdriven versus practitioner-facilitated approaches that use DNA-based diet information; and 2) qualitative investigation of participants' experiences to help interpret the intervention's quantitative outcomes.

5-xi) Report any prompts/reminders used Not applicable 5-xii) Describe any co-interventions (incl. training/support)
Not applicable 6a) CONSORT: Completely defined pre-specified primary and secondary outcome measures, including how and when they were assessed Nutrition-related Outcomes: All dietary intake data collection was done according to standard procedures [24,28]. Three-day food records measured pre/post caloric, macronutrient, micronutrient, and food group intakes which were compared to nutrition standards. The Canadian version of the Healthy Eating Index (HEI-C) [29] was used to assess diet quality. , standard measurements (detailed in measurements section), and protocols for nutrition assessment [25]. All questionnaires contained 12 or less screens (pages) and were pilot tested with study staff and student volunteers (n=11) to assess usability and technical functionality. Participants were emailed instructions and the links to each online questionnaire at the appropriate times during the study.

6a-ii) Describe whether and how "use" (including intensity of use/dosage) was defined/measured/monitored Not applicable 6a-iii) Describe whether, how, and when qualitative feedback from participants was obtained
Textual data from the online questionnaires (e.g., participant's personal dietary goals) were grouped into categories where feasible. Data from the focus groups were transcribed, entered into NVivo [44], and analyzed by research team members using interpretative thematic analysis to identify patterns, concepts, themes, and examples in relation to existing behavior change theories and the study objectives [45]. Interpretations were reviewed by research team members and participants to check for descriptive and interpretative validity. Qualitative data was reported based on thematic analysis derived from three independent reviews of the textual data. 6b) CONSORT: Any changes to trial outcomes after the trial commenced, with reasons Participants selected for the study included healthy medically stable adults (35-55 years) residing in the greater Vancouver area of the province of British Columbia, Canada. Participants were recruited via social media, a newspaper article, and posters. 7a) CONSORT: How sample size was determined 7a-i) Describe whether and how expected attrition was taken into account when calculating the sample size Detailed in protocol paper 7b) CONSORT: When applicable, explanation of any interim analyses and stopping guidelines Nutrition-related Outcomes: All dietary intake data collection was done according to standard procedures [24,28]. Three-day food records measured pre/post caloric, macronutrient, micronutrient, and food group intakes which were compared to nutrition standards. The Canadian version of the Healthy Eating Index (HEI-C) [29] was used to assess diet quality. Health Related Quality of Life (HRQOL) SF-8 (Short Form 8): A validated measurement tool of quality of life, functional health and well-being [32] based on a 4-week recall period. General Self E cacy (GSE): A validated item measure of self-e cacy shown to correlate with emotion, optimism and work satisfaction [34]. Measures of Change in Knowledge, Motivation, and Behavior: Three questions to assess for changes in knowledge, motivation, and behavior related to DNA-based dietary advice were developed by the authors based on the Stages of Change Model [36] and current review of the evidence. 8a) CONSORT: Method used to generate the random allocation sequence Not applicable 8b) CONSORT: Type of randomisation; details of any restriction (such as blocking and block size) Participants were randomized, using a random number generator by a statistician independent to the study into either the intervention or control group. 9) CONSORT: Mechanism used to implement the random allocation sequence (such as sequentially numbered containers), describing any steps taken to conceal the sequence until interventions were assigned Participants were randomized, using a random number generator by a statistician independent to the study into either the intervention or control group. 10) CONSORT: Who generated the random allocation sequence, who enrolled participants, and who assigned participants to interventions Statistician 11a) CONSORT: Blinding -If done, who was blinded after assignment to interventions (for example, participants, care providers, those assessing outcomes) and how 11a-i) Specify who was blinded, and who wasn't participants blinded 11a-ii) Discuss e.g., whether participants knew which intervention was the "intervention of interest" and which one was the "comparator" Wording was such that participants could not tell -both groups received dietary information 11b) CONSORT: If relevant, description of the similarity of interventions Not applicable 12a) CONSORT: Statistical methods used to compare groups for primary and secondary outcomes Food Intake and Nutrient Analysis: Nutrient analysis was conducted using ESHA -The Food Processor Nutrition Analysis and Fitness software and the Canadian Nutrient File [41,42]. Averages of the three days of nutrient values were used in the analysis. FFQ values were used to derive usual intakes of the nutrients of interest [24]. Descriptive and Inferential Analysis: Means (± standard deviations) or medians (and interquartile range) were reported based on a given continuous variable's distribution. Subject characteristics, group comparisons, and pre/post intervention differences were analyzed using Student t-tests, binomial tests of two proportions,Fischer exact tests, and two-way repeated measures ANOVA with Bonferroni's posthoc tests where appropriate. All analyses were done on an intent-to-treat basis usingSTATA software 12a-i) Imputation techniques to deal with attrition / missing values Outlined in previous question 12b) CONSORT: Methods for additional analyses, such as subgroup analyses and adjusted analyses Outlined in Q 12 a) RESULTS 13a) CONSORT: For each group, the numbers of participants who were randomly assigned, received intended treatment, and were analysed for the primary outcome 478 persons expressed interest in study participation. Ratio of females and males that were interested in the study was ~8:1, and although all males on the recruitment list were contacted there was imbalance in the male/female participant ratio. A total of 180 (38%) were invited to complete the online eligibility screening questionnaire; 73 of the invited individuals (73/180; 41%) were deemed eligible. Of the 73 eligible individuals, 58 enrolled and 55 (95%) completed the baseline health assessment questionnaire and food records (55/73; 75%). The final sample consisted of 55 adults between 37 and 57 years old (mean=45.8 years, SD±5.8).

13b) CONSORT: For each group, losses and exclusions after randomisation, together with reasons
As detailed in previous question; also a figure provided 13b-i) Attrition diagram Figure 2 14a) CONSORT: Dates defining the periods of recruitment and follow-up 3 months 14a-i) Indicate if critical "secular events" fell into the study period Not applicable 14b) CONSORT: Why the trial ended or was stopped (early) Not applicable Table 1 15-i) Report demographics associated with digital divide issues Not really applicable 16a) CONSORT: For each group, number of participants (denominator) included in each analysis and whether the analysis was by original assigned groups 16-i) Report multiple "denominators" and provide definitions Table 1 and 3 16-ii) Primary analysis should be intent-to-treat All analyses were done on an intent-to-treat basis using STATA software [43]. 17a) CONSORT: For each primary and secondary outcome, results for each group, and the estimated effect size and its precision (such as 95% confidence interval) Table 3 17a-i) Presentation of process outcomes such as metrics of use and intensity of use Table 3 17b) CONSORT: For binary outcomes, presentation of both absolute and relative effect sizes is recommended Table 3 18) CONSORT: Results of any other analyses performed, including subgroup analyses and adjusted analyses, distinguishing pre-specified from exploratory Changes in Dietary Intake, Anthropometrics, Self-E cacy and Quality of Life Signi cant pre/post intervention differences (Table 3)

19-i) Include privacy breaches, technical problems
Yes not applicable though 19-ii) Include qualitative feedback from participants or observations from staff/researchers Based on insights from our qualitative data, the variable findings among intakes of vitamin and mineral intakes may be due to a lack of understanding by the participant of their functions and relevance. DISCUSSION

20) CONSORT: Trial limitations, addressing sources of potential bias, imprecision, multiplicity of analyses 20-i) Typical limitations in ehealth trials
This study's strengths include its high retention rate, focus on a de ned adult population, assessment of dietary intakes using FFQ estimates and three-day food records to reduce misreporting error, and the provision of quantitative and qualitative data. However, this investigation could have been strengthened by including objective measures such as biochemical indicators of nutrient status. The modest sample size prevented strati cation of results based on individual genes. Furthermore, given the composition of the sample were mainly female and Caucasian the generalizability of the results are limited.

21) CONSORT: Generalisability (external validity, applicability) of the trial findings 21-i) Generalizability to other populations
This study's strengths include its high retention rate, focus on a de ned adult population, assessment of dietary intakes using FFQ estimates and three-day food records to reduce misreporting error, and the provision of quantitative and qualitative data. However, this investigation could have been strengthened by including objective measures such as biochemical indicators of nutrient status. The modest sample size prevented strati cation of results based on individual genes. Furthermore, given the composition of the sample were mainly female and Caucasian the generalizability of the results are limited.

21-ii) Discuss if there were elements in the RCT that would be different in a routine application setting
Providing participants with DNA information related to diet improved knowledge, motivation, and action related to healthy eating. However, tailored practitioner-led gene-based personalized nutrition interventions tend to be more effective in improving dietary intakes of key target nutrients such as fat and sodium. 22) CONSORT: Interpretation consistent with results, balancing benefits and harms, and considering other relevant evidence