Athletes’ optimal nutritional demand is dictated by their sport-related energy demand and training regimen as well as athlete’s own metabolic requirement. Maintaining optimal nutritional need would improve performance and recovery from exercise and injury whereas inadequate nutrition could compromise both. Such optimal nutritional need varies among different sports although the most remarkable finding when reviewing the literature is the scarcity of such data . Supplements are frequently used by athletes to compensate for nutritional deficiencies and boost nutritional consumption, aiming to achieve optimal energy demand. However, the effectiveness of supplements and their potential adverse effects remains questionable. Information regarding athletes’ supplement consumption is rather scarce and depends mostly on interviews and surveys.
To study the effect of specific dietary components on human health and performance, several groups have utilized metabolomics. Vitamin E supplementation was shown to influence phospholipid metabolism and induce lysoPC generation, a general pro-inflammatory response . A flavonoid-rich diet triggered changes in 63 plasma metabolites with 70% belonging to lipid and xenobiotic super pathways . Polyphenol soy protein complex supplementation was linked to an enhanced gut-derived phenolic signature and ketogenesis in runners during recovery from a 3-d heavy exertion . Consumption of fruits, such as banana and pears, was shown to improve 75-km cycling performance, reduce fatty acid utilization and oxidation, and enhance the anti-oxidant capacity by increasing unique phenolic production . Pistachio nut ingestion was also associated with improved 75-km cycling time and enhanced post-exercise plasma levels of raffinose, sucrose, and metabolites related to oxidative stress . Further studies have adopted a predictive metabolomics approach by examining the effect of composition of macronutrients consumed immediately post-exercise on the serum metabolic profile in the early recovery phase. These studies have suggested that pro-anabolic processes were favored, with a carbohydrate-protein mix compared to water or carbohydrate consumption . Despite their controlled nature, these studies are limited to the supplement in question and may not reflect what is truly consumed by elite athletes.
This study, therefore, followed a different approach by profiling supplement consumption in elite athletes retrospectively by identifying xenobiotics in their sera collected at two Anti-Doping Laboratories. A number of xenobiotics varied significantly among sport groups, including some that potentially originated from drugs, supplements, food products and other chemical contaminants. Although the exact sources of these metabolites and their potential effect on athletes’ performance remain to be confirmed, this study provides a snapshot of xenobiotics present in different sport groups that may carry certain benefits or inflict harm to these athletes.
Xenobiotics increased in athletics
Athletics that includes a group of competitive sports such as running, jumping, throwing, and walking showed higher concentrations of two xenobiotics that have potentially originated from food products and/or supplements, namely eugenol sulfate and stachydrine. Eugenol is a potent anti-oxidant found in many plants, herbs and spices especially in clove but is also contained in some supplements that claim blood purification and reduction of risk of gingivitis and heart disease . Stachydrine, a biomarker of citrus fruits consumption , was also increased in athletics. It is also contained in some supplements promoting calming and soothing effects and relief from anxiety. It can serve as an osmoprotective compound for the kidneys and has been shown recently to exert anti-inflammatory and anti-oxidative stress effects in animal models . There was a partial correlation revealed by GGM sub-networks between stachydrine and methyl glucopyranoside (alpha/beta), also elevated in athletics, suggesting a similar source, perhaps orange juice as previously shown . Other compounds elevated in athletics included 2, 3-Dihydroxy-isovalerate, a substrate of dihydroxyacid dehydratase known to be sensitive to nitric oxide  and 4-hydroxyhippurate, a microbial end-product derived from polyphenol metabolism by the microflora in the intestine.
Xenobiotics increased in football (soccer)
Footballers have shown higher levels of a number of xenobiotics that potentially originate from food products and/or supplements. These include caffeic acid (3,4-dihydroxycinnamic acid) that is produced by all plants including thyme, sage, spearmint, cinnamon, star anise, black chokeberry, tea and coffee and can also be consumed as a supplement. Caffeic acid-treated animals had enhanced exercise tolerance, lower blood lactate and hepatic oxidation . The derivative of caffeic acid (caffeic acid phenethyl ester) was shown previously to protect against hyperthermal stress induced by prolonged exercise . Caffeic acid has also anti-oxidant properties shown both in vitro and also in vivo [27, 28]. Ferulic acid 4-sulfate, a potent ubiquitous plant anti-oxidant found in high concentration in wheat, rice, peanuts, oranges and apples  and in oral supplement form, was also high among footballers. Ferulic acid is a ubiquitous plant component generated from phenylalanine and tyrosine metabolism and is a direct product of caffeic acid in plants. Other xenobiotics with higher levels in footballers included potential drugs such as ectoine, a nasal spray and eye drops, mostly used for the treatment of allergic rhinitis and rhinoconjunctivitis symptoms for relief of nose block and sneezing , potentially substantiated through continued grass exposure. Another potential drug-related compound is quinate. Quinine, an alkaloid that is synthesized in plants and the active ingredient of quinate, is a used for treating muscle cramps among footballers . Hippurate, a benzoate metabolite, was also higher among footballers. Hippurate is abundant in fruits and whole grains and has been shown to be associated with reduced risk of metabolic syndrome . From the GGM sub-networks, hippurate levels were partially correlated with other benzoate metabolites including catechol sulfate and O-methylcatechol sulfate, both shown to be also increased in football. Other xenobiotics elevated among footballers included 4-vinylguaiacol sulfate, a flavor additive in beer, also found in partial correlation with ferulic acid as shown by GGM sub-network. 2-furoyl glycine, partially correlated with quinate, was also high in footballers. 1.3.7-trimethylurate, a minor metabolite of caffeine, was also elevated in football.
Xenobiotics increased in boxing
The boxing group, albeit small (n = 17) and mainly composed of females, has shown three elevated xenobiotics in their sera compared to other sport groups. One of these was retinol (vitamin A), a nutritional supplement that has potent anti-oxidant properties and is used as an anti-aging cream . Another xenobiotic increased in boxing was 2.pyrrolidinone, the simplest γ-lactam with nootropic effects, providing neuroprotection after stroke and proving efficacious as an antiepileptic agent . The presence of this drug among boxers may reflect a prophylactic treatment for multiple head injury. Thioproline, an intracellular sulfhydril antioxidant and free radical scavenger that enhances immune functions, was also found to be elevated in boxers. Data from in vivo studies in mice showed that thioproline induces an anorexic effect associated with better survival and improved neurological function through a decreased oxidative damage .
Although this is the first study analyzing xenobiotics present in elite athletes from different sports, the lack of important information about participants such as age, body mass index and dietary and training regiments has hindered data interpretation. Additionally, the over representation of one group (footballers) compared to other participating groups may have introduced some bias in the study design that may have influenced the results. Furthermore, a batch effect may have also influenced the data, as athletes’ samples were collected and processed at multiple sites, although a batch correction was applied as described in the methods section. Another potential limitation of this study is the lack of information related to the precise role of each athlete in their team, which could impact their energy and endurance requirement . However, the large number of participants and their wide range of sport groups may have diluted out the effects of some of these potential confounders that were unaccounted for in our statistical model.
Assessment of the impact of changes in the dietary nutrient content on metabolic profiles is rather complicated as it overlaps with non-nutrient signals absorbed through environmental exposure. Both nutrient and non-nutrient contents are further processed by the gut’s microflora, thereby producing significant metabolic signals and adding to the complexity of the metabolome of biofluids in human nutrition . Therefore, recommendations on best practices when performing human intervention studies such as integration of OMICS data (including microbiome) with habitual dietary and lifestyle information (standardized FFQ) were recently suggested for better data interpretability [37, 38].