Abstract
Objective: Ingesting a mix of glucose and fructose during exercise increases exogenous carbohydrate oxidation while minimizing gastrointestinal (GI) distress. Several studies have suggested that a glucose-to-fructose ratio of 1.2:1 to 1:1 is optimal. No studies have quantified saccharides consumed during a nonsimulated endurance event. The aim of this investigation was to quantify saccharide sources used during an ultra-endurance triathlon and provide a resource for athletes desiring to manipulate the saccharide content of carbohydrate consumed during training and competition.
Methods: Participant self-report and direct measurement were used to assess foods and beverages consumed during an ultra-endurance (70.3-mile) triathlon. Manufacturer-supplied information, high-performance liquid chromatography, and the US Department of Agriculture Food Database were used to quantify saccharide profiles of foods and beverages. Participants reported GI distress during the run on a 0–10 scale. A subanalysis examined associations between saccharides and GI distress among participants consuming ≥ 50 g·h−1 of carbohydrate during the swim and cycle.
Results: Fifty-four participants (43 men) used 80 foods and beverages with a unique saccharide profile. Of total carbohydrate, median proportions as glucose, fructose, and sucrose were 64%, 5%, and 10%, and only 7 foods (8.8%) had a glucose-to-fructose ratio of 1.2:1 to 1:1. The median glucose-to-fructose ratio of carbohydrate ingested was 2.9:1 (2.2:1–5.3:1). Twenty participants consumed ≥ 50 g·h−1 of carbohydrate during the swim and cycle, and significant correlations with incident GI distress at mile 1 of the run were found for glucose (r = 0.480, p = 0.032) and fructose (r = −0.454, p = 0.044).
Conclusions: The majority of foods and beverages consumed during an ultra-endurance triathlon did not contain an optimal saccharide profile. Furthermore, glucose intake was associated with greater GI distress among participants consuming a high rate of carbohydrate.
ACKNOWLEDGMENTS
Morgan Betker, John Fitzgerald, Christopher Lundstrom, and Zachary Rourk deserve acknowledgement for helping with data collection. We thank DuTriRun for accommodating the study and helping with recruitment.
FUNDING
This work was supported in part by the University of Minnesota Doctoral Dissertation and Hauge Fellowships.