Investigation of in Vivo Endogenous and/or Exogenous Production of Phenolic Metabolites Using (un)Targeted Metabolomics

The study will consist of a randomized cross-over trial and each arm will last 5 days. A total of thirty (30) healthy volunteers will be recruited for the study in Parma (PR), using announcements placed in the university, hospital, and public places. To allocate subjects to each arm, a list will be generated using a randomized block design by using Random Number Generator Pro (Segobit Software). This list will be blind for the PI and volunteers, as it will be made by a colleague not involved in the subjects' enrolment and using a numbered sequence in sealed, opaque envelopes.

Volunteers freely accepting to participate to the study will be asked for their written informed consent. Information to the volunteers will be provided before and separately from the consent form. These information sheets and the informed consent forms have been prepared according to the Oviedo Convention and consider the guide of the local Ethics Committee. After obtaining their written informed consent, volunteers will undergo a screening visit to determine eligibility and collect personal data.

Privacy of the volunteers will be assured through pseudoanonymization by the attribution of a personal alphanumeric code. Recruited volunteers will also sign an authorisation to the use of personal information and retrieved data.

After recruitment, subjects will attend a first visit, when personal data will be collected, and body composition and anthropometry parameters measured. Then, they will receive the foods, supplements, and materials necessary to follow the study. The recruited subjects will be requested to follow a low-(poly)phenol diet, which will be the same for all the volunteers and maintained for the whole period of the intervention study. The foods, preparation mode, and times for eating the meals will be provided by the sponsor. The diet will be personalised to approximate the necessary calories based on the measured basal metabolic rate, multiplied by the physical activity level set at 1.45, and there will be quantitative variations on the low-(poly)phenol diet based on the daily energy requirement: 1500 kcal/day, 1750 kcal/day, 2000 kcal/day, 2250 kcal/day, 2500 kcal/day, 2750 kcal/day, or 3000 kcal/day, all with nutritionally balanced amounts of protein, fat, and carbohydrates. In addition, participants will be requested to restrict physical exercise during the study duration.

Subjects will be asked to collect urine produced during the 60h prior the intervention, starting from the evening prior the first day of diet (from 21.00 of the night prior the first day of diet) and every 12 hours (60-48h, 48-36h, 36-24h, 24-12h, 12-0h), and to complete the food diary to check for compliance to the diet. Subjects are asked to bring collected urine to the the Human Nutrition Unit of the Food and Drug Department of the University of Parma.

On the morning of the third day of diet, subjects will be requested to arrive at Parma University in fasting conditions. Prior to receiving the intervention, fasting spot urine and saliva samples will be collected. Then, sugared-decaffeinated-coffee (3 doses of 50 mL each and 5g of sugar) or sugared-water (150 mL of water and 5g of sugar) will be provided according to randomization of the arms, along with a standardized breakfast consisting of a (poly)phenol-free, nutritionally balanced solid meal and water (ad libitum).

After breakfast, subjects will be asked to autonomously collect urine for the next 48h at specified timepoints (0 - 3h; 3 - 6h; 6 - 9h; 9 - 12h; 12 - 24h; 24 - 36h; and 36 - 48h), following the same personalized diet. Lastly, subjects will autonomously collect a fecal sample within 6 to 48h after the intervention.

On the morning of the last day, after the last urine collection, subjects will bring collected urine and fecal samples and the 5-day food diary to the Human Nutrition Unit of the Food and Drug Department of the University of Parma.

After the first arm, subjects will resume their habitual lifestyles for 2 consecutive weeks, subsequently, they will attend another visit, when all the anthropometric evaluations carried out during the first visit will be repeated and another kit with foods and information for the personalized diet will be given. With the study being a cross-over design, the protocol followed will be the same as the one described above (except for saliva sampling, which will only be performed once). Coming the third day of diet, subjects who ingested coffee in the prior arm of the study will ingest water, and vice-versa.

For this study, the sample size was estimated considering the mean concentration of coffee-derived (poly)phenols and their metabolites in urine as the primary outcome and the available literature on similar approaches and studies of the area. The number of 30 volunteers was established considering the accepted parameters for clinical trials (power set at 0.8 and α set at 0.05) possible dropouts (15% drop-out rate), and exclusion of volunteers to maintain the sample power for statistical analysis. Before any comparison is performed, the normality of each variable will be evaluated to choose the most appropriate statistical test. If metabolite data are normally distributed, data will be expressed as mean ± standard deviation and analysed using repeated measures t-test or ANOVA with post-hoc comparisons. If data are not normally distributed, they will be reported as median and Interquartile Range and the Friedman test with post hoc pairwise comparisons will be performed. P-value < 0.05 will be regarded as statistically significant.

Unsupervised and supervised multivariate analysis will be used to identify and characterize the different urinary metabotypes, following the procedures developed by our research group for targeted metabolomics analysis of phenolic metabolites or when using untargeted metabolomics data. We will be using different clustering approaches to better explore the existing variability. This will serve to establish the number of clusters (i.e., phenolic metabotypes), the number of participants forming each cluster and the similarities among possible metabotypes.

The relationship between patterns of LMWP and their putative determinants (age, sex, ethnicity, lifestyle, anthropometrics, body composition, genotype, and gut microbiome) will be explored using correlation tests and multivariate analyses. P-value < 0.05 will be regarded as statistically significant. Statistics will be carried out by the specialized staff of the Human Nutrition Unit, under the supervision of the PI, and in collaboration with the University teams in charge of microbial profiling and genotyping.