Despite the accumulative scientific evidence supporting the interest in eating plant food bioactives to protect essential body functions and reduce the risk of developing cardiometabolic diseases, the human intervention studies aiming to demonstrate the effects of specific phytochemicals, or foods rich in, on intermediate biomarkers of cardiometabolic risk have often shown mixed results [
30]. One major cause is the heterogeneity of the individual response to their intake, which reduced the significance of the effects on some biomarkers of health at the scale of the population studied, although their intake showed promising results in subgroups of subjects participating in the trials [
31]. This between-subject variation in the response suggests that the consumption of particular foods or bioactive compounds may benefit some individuals more than others. However, up to now in published controlled trials, the IIV has been most often masked by the statistical analysis which produces the mean data of the study population rather than the individual’s data [
32]. To overcome this issue, in future human trials, the data obtained for each subject should be published as those provide very informative data on the variation of the effect among the individuals. If the human trial is complemented with a study of the variation on food bioactive compound ADME, and with a measurement of the exposure to specific bioactive metabolites, a correlation between the biological effect on cardiometabolic health biomarkers, and the exposure to the bioactive metabolites could be evaluated. This approach will likely explain in large part the differences observed between responders and non-responders to the intervention and will allow identifying the metabolites responsible for the bioactivity in the body of the different categories of plant food bioactives. This heterogeneity in responsiveness has often led to inconclusive results in clinical trials examining the health effects of specific phytochemicals over the last decades, with the unfortunate underestimation of the actual role of PBF in the health benefits of plant-based foods being a possible consequence. A full consideration of PFB in the future strategy of personalised nutrition cannot escape from a first in-depth investigation of the factors responsible for the IIV in response to consumption of the major categories of PFB of our diet.
A range of factors such as genetic background, gut microbiota composition, pathophysiological status, age, or gender could explain these interindividual variations, and they may differ depending on the bioactive compounds. For example, the IIV in the bioavailability of carotenoids has been shown to depend on single nucleotide polymorphisms (SNPs) in genes involved in the intestinal uptake or efflux of these compounds as well as in genes involved in their metabolism and transport [
7]. The phenotypic effect of each SNP is usually low, but combinations of SNPs can explain a significant part of the variability with impact on the carotenoid status of individuals. Variability in polyphenols ADME can also partly originate from differences in gut microbiota composition which determine the conversion or not of some categories of (poly)phenols into metabolites displaying higher bioactivity than the parent compounds and with potential impact on their health effects [
33]. For example, different (poly)phenol metabotypes (reflecting the metabolic capacity of the gut microbiota towards dietary (poly)phenols) have been described for the gut microbial metabolism of soy isoflavones and ellagitannins from berries and nuts, depending on their conversion or not into equol and urolithins respectively [
34,
35]. Clustering of individuals according to equol or urolithin metabotypes is consistent with the IIV observed in the improvement of cardiometabolic biomarkers in human intervention studies using isoflavones or ellagitannins [
36,
37]. However, between-subject variations in the biological response to (poly)phenol consumption are not always linked to differences in bioavailability. For example, as observed in a parallel-group intervention study, at comparable levels of bioavailability, the lowering effect of cocoa flavanols on systolic blood pressure reached significance in young but not in elderly subjects [
38]. Acute intake of curcumin has also shown to improve endothelial function in women only and not in men, whereas the plasma concentrations of curcumin were unchanged between the males and females enrolled in the trial [
39]. Substantial heterogeneity in the individual LDL cholesterol response to phytosterol therapy has also been repeatedly reported in human interventions [
40], and the variability has been mainly explained by polymorphisms across genes associated with cholesterol trafficking pathways [
41]. Despite these few known examples, the factors responsible for variability in ADME and biological responsiveness for the main categories of PFB are far from being fully identified. The available studies do not allow to determine to what extent the between-subject variation in the ADME of PFB can explain the IIV in their biological responsiveness due to the lack of clinical trials investigating both the bioavailability and the biological effects [
32].
Although some examples in the literature clearly illustrate that IIV in response to PFB consumption exists, only limited and scattered data are available on the subject. To tackle the complexity of this question through a multidimensional approach, the European COST POSITIVe network (
https://www6.inra.fr/cost-positive) has gathered a large community of experts in nutrition and plant food bioactives, food science, clinical research, microbiology, gut microbiome, genetics, nutrigenomics, bioinformatics, cellular and molecular biology. These experts have joined their efforts to analyse systematically the currently fragmented knowledge in the field. The activities of this network aimed to (1) evaluate the extent of the IIV for the major categories of PFB and identify the main factors responsible for between-subject variation in both the ADME and the biological responsiveness regarding cardiometabolic endpoints; (2) integrate the main findings of the network to identify the gaps in knowledge and needs for future research and also to examine how this knowledge translates into concrete applications for the different categories of stakeholders; (3) provide some recommendations to better capture interindividual variation in intervention trials. The main results of the POSITIVe network are presented and discussed in the reviews of the special issue of the European Journal of Nutrition on interindividual variability in response to PFB.