Our research provides insight into voluntary pre-exercise food restrictions endurance runners use to mitigate GI symptoms, via a fully powered, reliable and validity tested questionnaire. Furthermore, we controlled for GI disorders, allergies, and food intolerances. Finally, we have representation from a diverse group of runners, allowing for increased specificity in our understanding of food avoidances.
Pre-exercise food restrictions
Foods most commonly avoided were milk products, high protein, high-fiber, chocolate, and caffeinated beverages, aligning with recommendations to limit protein, fat, and fiber [15,16,17] while concurrently highlighting the need for more in-depth research. Importantly, each food item is a complex mix of macro and micronutrients as well as other bioactive ingredients that can impact GI symptoms; thus, caution is advised when making conclusions regarding an individual nutrient in a food. Generally, the higher the performance level, the more likely an athlete was to restrict food, which is likely a function of increased GI symptoms with increasing exercise intensity [8, 25]. Another possibility is that higher level athletes are more experienced and therefore more familiar with aggravating foods.
Avoidance of milk and milk products was common among runners in our study. Interestingly, others have found that in female runners declaring they have an “outstanding diet”, intake of dairy beverages was significantly lower than those who rated their diet as “average” . Dairy products are complex foods and naturally contain lactose. We are in agreement with others who have demonstrated that athletes will remove sources of lactose; a high FODMAP food, from their diet to improve GI symptoms . Not all dairy products contain lactose, therefore, future research should subdivide this category into lactose containing and lactose-free products. Considering that lactose-free milk was also among the top foods avoided, the reasoning is likely multifaceted. Dairy products are also sources of fat and protein, which are thought to promote exercise-induced GI symptoms . Further analysis indicated that young females were most likely to avoid milk products followed by young males, which is in agreement with other findings indicating age as a negative predictor for dairy consumption, in the general population . Our research suggests that age remains a predictor for dairy avoidance in sport, even when allergies and intolerances are considered. Further, Yantcheva et al.  report the perception of mucous production as a common reason for the avoidance of dairy, which may be related to reports of phlegm or mucus as an “other” symptom. The role of dairy in mucus production has not been fully elucidated ; however, it clearly remains a popular perception.
Foods typically considered high protein, especially animal protein, were commonly avoided pre-exercise. Athletes are advised to avoid excessive protein pre-exercise ; however, recommendations for the appropriate amount of protein and studies regarding the effects of protein intake on exercise-induced GI symptoms in runners are lacking. Tiller et al.  recommend 1.6 g/kg body weight per day as a minimum for ultra-endurance runners; however, they do not specifically address the pre-race period. They do note that protein intake during specifically ultra-endurance running may positively affect energy metabolism and mitigate muscle damage with the caveat that the results are equivocal. Snipe et al.  explored 14.8 g whey protein intake pre/during a 2 h run, designed to induce exertional heat stress, as compared to water or 15 g glucose. A reduction in intestinal epithelial injury and intestinal permeability was found with both whey protein and glucose; however, gut discomfort and gastrointestinal symptoms were higher with protein. In basketball players, the addition of protein to the pre-exercise meal at 1 g/kg body weight resulted in increased gastrointestinal symptoms as compared to carbohydrate alone . The aforementioned studies support our conclusions regarding protein rich foods and increased exercise-induced GI symptoms. Given the potential performance benefits of protein in the pre-exercise meal and popularity of high-protein diets, additional research is required to determine if there is a threshold below which symptoms are minimal. Generally, the higher the performance level, the more likely athletes were to avoid high protein foods, possibly due to increased GI symptoms with increased exercise intensity [8, 25].
Runners avoided high-fiber foods, which aligns with others describing lower intakes of dietary fiber by endurance runners, although not specifically in the pre-exercise meal [32, 33]. Limiting of dietary fiber pre-exercise is advised [16, 17], based largely on a study by Rehrer et al. , linking dietary fibers to intestinal cramps. Dietary fiber ingestion is associated with decreased splanchnic vascular resistance resulting in increased splanchnic vasodilation and splanchnic flow. These physiological effects oppose blood flow needs during exercise, where there is prolonged splanchnic hypoperfusion, and consequently may present as abdominal disturbance . Delayed gastric emptying with viscous dietary fibers  may also play a role, further exacerbated by high intensity exercise . Conversely, insoluble fibers stimulate peristalsis via fecal bulking , which may promote urge to defecate or diarrhea. Fiber restriction was more common in longer distances, which may reflect changes in GI transit time, as diarrhea was reported more frequently in longer distances. Finally, many fiber rich foods are high FODMAP foods, which may provide another mechanistic explanation. Dietary fibers are heterogeneous and vary in their physiological effects; thus, recommendations should consider fiber type in addition to quantity. Further, given the health benefits of adequate fiber and recent advances in the understanding of the importance of the gut microbiota in athletes , research is required regarding pre-exercise timing strategies to optimize intakes while minimizing GI symptoms.
Dark chocolate has been proposed as an ergogenic aid via increased nitric oxide ; however, caution should be advised given our results of high avoidance pre-race, although we did not distinguish between the different types of chocolate. Chocolate has been described as a food item that provokes GI disturbance, particularly constipation , although research in athletes is lacking. Chocolate contains several biologically active compounds including cocoa, caffeine, and fat; thus, the mechanisms are unclear. It is known, however, that high fat foods may aggravate exercise-induced GI symptoms .
Coffee and tea represent another food group often avoided and morning caffeine intake has been associated with increased GI symptoms in the lower gut of triathletes . Interestingly, competitive athletes and longer distance runners were less likely to avoid coffee/tea and this may be a reflection of the potential ergogenic effects of caffeine in endurance exercise . Future research should consider the effects of coffee, tea, and herbal infusions separately as they contain different constituents and can result in different physiological effects.
Energy beverages are a cocktail of vitamins, sugars, and plant extracts, especially stimulants. GI upset is included in the list of commonly reported symptoms after energy drink use  and safety is a concern . Energy drinks were restricted more often in younger athletes; however, they are a relatively new product and marketing of energy drinks is typically youth oriented. It is possible that older athletes would not report avoiding energy drinks if they were unfamiliar with the product.
Exercise-induced gastrointestinal symptoms
GI disturbance during runs is a common concern as described here and throughout the literature [1, 7,8,9]. At a minimum, GI symptoms associated with exercise are related to mechanical forces, altered GI blood flow, changes in the GI mucosal activity, neuroendocrine changes, and stress [1, 4, 7,8,9].
Female runners were more likely to experience urge to defecate and diarrhea. Additionally, young females reported highest rates of gas, nausea, fullness, and stomach pain/cramps, which supports research examining GI symptoms in an exclusively female running cohort where younger age was related to increased GI symptoms . Further, others support a higher prevalence of GI symptoms in female athletes [3, 22, 45, 46]. Conversely, in a trial to determine the effect of biological sex on GI symptoms during exertional-heat stress, by timing testing during the follicular phase of the menstrual cycle, no differences in GI symptoms were reported except for flatulence and abdominal stitch, which were higher in males . The aforementioned findings suggest further research is required to determine the causes of increased symptoms in females and the potential relationship to sex hormones and female gut physiology.
The higher prevalence of symptoms in younger athletes is confirmed by the literature [44,45,46]. Increased age may protect against GI symptoms due to diminished splanchnic vasoconstriction via reduced catecholamine response and consequently increased oxygen supply . Further, increased age often reflects increased running experience, which is associated with fewer GI symptoms [44, 45].
Considering performance level, it was often the lower recreational athletes who were least likely to report symptoms. Potentially these athletes are competing at lower intensities, thus have fewer symptoms, as GI symptoms are reported to increase with exercise intensity [8, 25]. Symptoms are thought increase with distance , however, studies are required, and are likely compromised by the tendency to consume food and fluid during the longer events. In our study, diarrhea increased with the longest distances run after controlling for other factors.
When the reasons for food avoidances were explored, the majority of the participants relied on personal experience or personal preference. Further investigation into their sources of information and how this varies by age, gender, event, and performance level is of interest.
A limitation to the study is its observational nature, which precludes any causal conclusions. Conversely, the study does highlight candidate foods for future clinical trials, as it is not feasible to test every food in a controlled study. The potential confound of a food intolerance was considered by removing those individuals with reported food intolerances; however, this was not always clear for combination foods such as smoothies or high-fiber foods. Fortunately, food intolerances were typically clearly identified. With respect to food categories: fats, oils, spicy foods, and high FODMAP foods should be added to future questionnaires. An “other” section, where people reported avoiding high fat and spicy foods, was included, suggesting these are areas of concern. Additionally, we did not ask participants to indicate the severity of their symptoms or provide a symptom for each food avoidance; thus, we cannot associate a specific food to a specific symptom or comment on the degree of discomfort. Finally, multiple comparisons were made with a 5% level of significance; thus, there is a risk of false positives.