Paper Selection Leads to a Misleading Conclusion: Updated Evidence of Ice Slurry Ingestion on Endurance Performance

This short report added to the scientifi c debate regarding a controversial conclusion published in European Journal of Sport Science. Reasons for the confl icting views have been presented in detail. Importantly, updated evidence suggests that ingestion of ice slurry is an eff ective countermeasure for endurance performance in the heat. Endurance athletes competing in the upcoming Tokyo 2020 Summer Olympics are encouraged to continue utilizing this simple yet very eff ective method for best possible performance.


Introduction
In a paper published in European Journal of Sport Science in 2018, Choo, Nosaka, Peiff er, Ihsan, and Abbiss (2018) concluded that ice slurry ingestion had no clear eff ect (Hedges' g, 0.2; 95% confi dence interval, −0.07 to 0.46) on endurance performance. I am not convinced of the validity of their literature selection and have great concern about what has been conveyed to those readers relying on science to guide their practice in the fi eld.

Methods
First, the authors' conclusion was biased by one particular study (Byrne, Owen, Cosnefroy, & Lee, 2011) which was incorrectly included in the meta-analysis. Th e methods of that paper were described in detail in the original article, "Cold fl uid was prepared by mixing refrigerated fl uid (approximately 4°C) with ice cubes in a vacuum fl ask to bring the temperature to 2°C". Mixing water with ice cubes is a common practice in sport science and in the fi eld to reach cold fl uid temperature. Th is however does not equate to ice slurry ingestion and the original article clearly stated that, "Th e cold fl uid in the insulated cup contained no ice". Th e internal cooling eff ect is diff erent between cold fl uid and ice slurry due to the enthalpy of the melting of ice. Choo and colleagues (2018) improperly included this study in their meta-analysis despite their purpose was to investigate the performance eff ect of ice slurry ingestion.
Second, I challenge Choo and colleagues' (2018) selection in their meta-analysis of another three studies (Gerrett, Jackson, Yates, & Th omas, 2017;James, Richardson, Watt, Gibson, & Maxwell, 2015;Zimmermann & Landers, 2015) as valid measures of endurance performance. When it comes to the measurement of elite endurance performance, a time trial off ers direct assessment and time to exhaustion as well provides reliable prediction of endurance capacity (Amann, Hopkins, & Marcora, 2008). Gerrett et al. (2017) measured eff ect of ice slurry ingestion on self-paced intermittent exercise, the test of which was originally designed for soccer-specifi c intermittent movements primarily taxing the anaerobic capacity. Th is study clearly did not assess the specifi c construct of endurance capacity for the purpose of their meta-analysis. Th e study by James et al. (2015)  between running and cycling are beyond the scope of this letter, sensitivity and reliability issues are of important consideration in high performance measurements. One important question needs to be answered to consider its validity: Has the original cycling-based lactate test been cross validated in the running mode? Again, repeated sprint ability is more related to short sprint ability than endurance ability and power-based repeated sprint test provides poor prediction of time trial (Balmer, Davison, & Bird, 2000). Th us, the test results from Zimmermann and Landers (2015) does not relate to the purpose of their meta-analysis. Notably, the aforementioned four studies all reported nonsignifi cant performance eff ects (Choo et al., 2018). Taken together, four out of eleven meta-analyzed studies were poor selection of the relevant literature, which biased the main results. Th erefore, the stated hypothesis could not be, and was not, answered with the papers selected for their review.

Results
To address the conclusion from Choo and colleagues' (2018) work, an updated meta-analysis has been performed excluding the four studies in question (Zhang, 2019) and reported a signifi cant eff ect size of ice slurry ingestion: Hedges' g, 0.60; 95% confi dence interval, 0.34-0.87. When this eff ect size is translated to performance eff ect, it represents 8.73% faster performance in the heat (Figure 1). Guy and colleagues (2015) have reported that elite endurance performance is aff ected in the heat (Figure 1).

Figure 1.
Change in endurance performance in the heat. Ice slurry ingestion resulted in 8.73% (weighted mean, by a random-eff ects meta-analysis) performance enhancement. Data are replicated from Zhang (2019). For comparison, elite distance performance in the heat was slower (unweighted mean) in hot environments. Data are replicated from Guy, Deakin, Edwards, Miller, and Pyne (2015).
Th us, ice slurry ingestion could eff ectively neutralize the negative eff ect of environment temperature on endurance performance and this is supported by the fi eld adoption among elite track and fi eld athletes (Periard et al., 2017).

Discussion
In less than two years, the Summer Olympics and Paralympics will be held in a very hot and humid Asian summer weather, which poses a real challenge to athletes' health and performance. Th e message from this letter to the practitioners is that, ingestion of ice slurry is eff ective in enhancing endurance performance and should be recommended for endurance events in the heat.