Biomarker pAtterns Reflecting habITual fOod iNtakE: the BARITONE Project

PURPOSE AND AIMS

The aims of the present project are to:

1. Identify and validate biomarker patterns of intake of the different food sources for protein, i.e., meat, fish and plant-based sources
2. Identify and validate biomarker patterns of high and low intake of the food group vegetables, fruit and berries.

Validation here includes evaluation of plausibility, robustness, and dose-response.

PROJECT DESCRIPTION Validated biomarkers are still lacking for food groups that are crucial when evaluating dietary intake, such as meat, fish and plant-based sources of protein (i.e., legumes such as lentils, peas and beans and cereal proteins such as oat- and wheat-protein); and for berries, fruits, and vegetables. For sources of protein, a well-controlled cross-over trial is needed where biomarkers of any intake of these different sources are evaluated in different biofluids. For berries, fruit and vegetables, a challenge is that no obvious control food group exists. Individuals with low consumption from this food group may instead consume any type of food. Hence, the investigators aim to evaluate biomarkers that separate consumers of low vs high intake. The investigators will follow the criteria for discovering and validating biomarkers suggested by Dragstedt et al (2018), here applied to patterns of biomarkers. The investigators will identify patterns of putative biomarkers of recent intake in biofluids in a well-controlled feeding trial (BARITONE-intervention) and develop prediction models of being a consumer/non-consumer or high/low consumer using machine learning. The robustness and dose-response of the models will be tested in data on habitual intake (BARITONE sample). Data for protein sources and berries, fruits and vegetables will be generated in the planned BARITONE-intervention trial.

Biomarker pAtterns Reflecting habITual fOod iNtakE: the BARITONE project The BARITONE project consists of two parts: 1) the BARITONE-sample (n=210) that undergoes a comprehensive baseline screening with extensive phenotyping including a 7-day weighed food diary (wk 1), and 2) the BARITONE cross-over intervention trial (n=70) where a selected group from the BARITONE-sample undergoes three dietary intervention weeks with two 7-day weighed food diaries during the two washout weeks in between.

The BARITONE-sample will be used to evaluate robustness and dose-response of putative biomarker patterns of habitual intake, because it will contain highly detailed data on habitual dietary intake and individual phenotype/metabotype. The intervention weeks of the BARITONE-trial will be used to identify putative biomarker patterns of recent intake of protein sources. Because it is a cross-over trial it allows for within-individual comparisons on metabolite reactions to dietary exposure. The wash-out periods between intervention weeks will be used to identify putative biomarker patterns of recent high or low intake of berries, fruits and vegetables, because participants will be selected based on baseline intake levels of these foods and here asked to return to their habitual diet. Hence, broad variation in intake should be represented.

The BARITONE sample for habitual diet In total, 210 healthy individuals aged 18-65 years will be invited through advertisements in media to detailed phenotyping at the laboratory of Clinical Nutrition, Gothenburg University. Procedures include a socio-demographic and lifestyle questionnaire (chronic illnesses, pharmacological treatment, physical activity, questions about menopause/timing in the menstrual cycle), a short food frequency questionnaire (FFQ), a 7-day weighed food diary and measurements of anthropometry (weight, height, body mass index, body composition by dual energy X-ray absorptiometry (DXA) or bioelectrical impedance analysis (BIA)). Biospecimens (fasting blood, spot urine; baseline stool samples) will be collected and analyzed for serum lipids, Hb, HbA1c, TSH, metabolomics including lipidomics and baseline gut microbiota.

The BARITONE cross-over intervention trial for recent intake of protein sources (Aim 1)

The BARITONE trial aims to identify metabolite patterns that can separate individuals with respect to intake of meat, fish or plant-based source of protein. From the BARITONE -sample, the investigators will select 35 participants with low intake of berries, fruit and vegetables and 35 with high intake of these same foods, to ensure a wide range of habitual intake of these foods to address Aim 2. Selection will be based on their baseline FFQ. The limit for low (<100g/day) and high intake (>400g/day) are based on the 15:th and 85:th percentile of intake data on vegetables, fruit and berries, from Riksmaten. Participants will be randomized by excel randomization function to order of trial diets by an external researcher. Blinding of participants is not possible in this diet intervention.

Intervention week diets: The diets will be identical in foods and macronutrient composition except for meat, fish or plant-based source of protein. Participants will be provided with all their food for the three trial weeks, with one week wash-out periods between the intervention periods when habitual consumption patterns should resume. Food bags will be delivered free of charge at the day before each intervention week including instructions, recipes and menus. Water will be the suggested drink. Non-sweetened drinks, coffee and tea will not be provided, but participants are allowed to consume these freely and report the intake and they should keep the consumption constant during all three intervention weeks. Alcohol will not be allowed during the intervention weeks.

The BARITONE wash-out weeks for habitual intake of berries, fruits, and vegetables (Aim 2): Detailed 7-days weighed food diaries will be collected during the two wash-out weeks among the trial participants, already selected to have low or high habitual intake of berries, fruits and vegetables. The diaries will be used to verify classification of the participants into high and low consumers. Metabolite patterns in the biofluids collected at the end of each wash-out period will be used to construct models for predicting intake of berries, fruit and vegetables, and thereafter robustness of the models will be evaluated in the BARITONE-sample.