The purpose of this study was to determine the influence of a 3-day, high- or low-carbohydrate diet on swimming economy in recreationally-trained swimmers. It was hypothesized that the HCLF diet would increase carbohydrate utilization relative to a LCHF diet, resulting in an increase energy conversion per volume of O2 consumed. Although RER was greater during exercise following three days of a HCLF diet, no differences in VO2 or Cs were detected. Therefore, these data do not support our original hypothesis because there was no improvement in swimming economy (i.e., a reduction in Cs) with the HCLF diet. In addition, HR did not differ between diets, although O2 pulse, a non-invasive estimate of stroke volume, was greater following the HCLF diet. To our knowledge, this is the first study investigating the effects of diet on swimming economy.
Results from the present study are in contrast with previous research reporting an effect of diet on movement economy in cyclists, runners, and race walkers (15–17). It is possible that exercise intensities used in our study may have been lower than in the previous studies and, therefore, more reliant on fat oxidation. In support of this, Shaw et al. (16) reported that running economy was impaired at intensities over 70% VO2max, but preserved at intensities lower than 60% VO2max, following 31 days of a low-carbohydrate diet in trained runners. Additionally, four weeks of a LCHF diet had no impact on cycling economy in endurance-trained athletes cycling at ~ 63% VO2max (22). Other research has shown that a LCHF diet impaired movement economy during exercise at 70–90% VO2max following 3-week diet and training interventions in elite male race walkers (17) and in recreationally-trained male runners (23). Additionally, three days of a high-carbohydrate diet (70% of energy intake) increased cycling gross efficiency (i.e., improved cycling economy) at 70–75% VO2max compared with both low- and moderate-carbohydrate diets in trained cyclists (15).
Differences in training status may also have affected our results, as higher-caliber athletes generally have better running economy (2) and cycling efficiency (24). The VO2max values of our participants averaged 42.4 ml kg− 1 min− 1 whereas other studies showing an effect of diet on economy used participants with a mean VO2max of 56–66 ml kg− 1 min− 1 (15–17). However, swimmers tend to have lower VO2max values measured while swimming compared to running (25, 26) and triathletes record higher VO2max values when cycling or running compared to swimming (26, 27), so we likely underestimated the true VO2max of our subjects. Previous research has also shown that 5–7 days of a LCHF diet reduced cycling efficiency in sedentary subjects, but not in trained athletes, possibly due to lower levels of uncoupling proteins found in the trained athletes (28, 29). Therefore, it is possible that differences in subject training status underpin the lack of differences observed between diets in our study.
Differences in mode of exercise may also explain the lack of effect of diet in the present study. For swimming, minimizing drag force and maximizing propelling efficiency (i.e., maximizing useful power and minimizing wasted power) are adaptations of higher-caliber swimmers that improves swimming economy (30). In addition, swimming technique likely plays a greater role in swimming economy than cycling or running technique do for economy in those sports as the swimmer moves through water. Therefore, any diet effects on swimming economy may have been masked by slight changes in external factors between trials.
There is also the potential that anaerobic metabolism was greater following the HCLF diet that could explain the increased economy as measured by indirect calorimetry. We did not assess blood lactate at the end of each submaximal swimming trial. However, Cole et al. (15), Burke et al. (17), and Shaw et al. (16) all measured blood lactate concentration and found no differences between diets during submaximal exercise intensities. In the present study, we found VO2 stabilized between 2–3 min into each 5-min stage (data not shown), indicating subjects were below their critical swimming velocity, above which VO2 and lactate will not stabilize (31).
Aggregate CS in our study ranged from 649–755 J m− 1. Higher CS values (690–1310 J m− 1) have been reported while swimming at faster speeds (32), while lower values (593 J m− 1) have been reported when swimming with legs held together by an ankle strap and supported by a pull buoy (33). In comparison, the energetic cost of running and race walking are approximately half of CS values reported in the present study (16, 17). Therefore, the values for CS are consistent with previous research, and swimming is more costly than running or race waking.
Oxygen pulse was higher following the HCLF diet. Although body weight and hydration status were not measured in this study, it is likely that the high-carbohydrate group had an increase in total body water when performing the swimming test due to increases in muscle glycogen (34). However, water stored with muscle glycogen would increase intracellular but not extracellular water content (35), though it is possible that this extra water may have played a role in the increased oxygen pulse observed during submaximal exercise in following the HCLF diet.
A strength of this study is that the diets were individually tailored for each participant, based on their habitual food choices. Participants were required to shop for and prepare all of their own food in the quantities specified, and could communicate with the researchers about food choices and potential substitutions if needed. This design increases the ecological validity of this study as most recreationally-active swimmers do not have access to the same resources as collegiate and professional swimmers whose diet and exercise volume can be more closely monitored. It is possible that participants could have been untruthful in recording their diets, although the difference in RER between diets suggests acceptable adherence. Protein intake was significantly different between groups (16.0 for HCLF vs. 17.6% for LCHF), but this difference is likely too small to impact the results of this study.