Reply on Baker’s comments to Nolte and Noakes: “change in body mass accurately and reliably predicts change in body water after endurance exercise”

Abstract

The title of the original article by Dr Baker and her colleagues was: ‘‘Change in body mass accurately and reliably predicts change in body water after endurance exercise’’. In her response to our letter Dr Baker reduces her certainty. She now concludes only that the change in body mass is a ‘‘reasonable method to estimate hydration status’’. This is a more reasonable conclusion that better reflects the published literature. But the focus of our letter was not to show that change in body mass (DBM) is an unreliable method for estimating hydration status. Why would we try to disprove a relationship that we believe exists (Nolte et al. 2010a, b; Tam et al. 2009)? There were two reasons for our letter. First, we concluded that the methods used by Baker et al. (2009) to measure total body water (TBW) may have been relatively imprecise since their study found a significant relationship between DBM and the change in TBW (DTBW) in a total of 62 measurements only when those data were analysed in a manner that, for the reasons we described in detail, we consider inappropriate. The potential methodological errors we identified might explain why their data did not show a relationship between DBM and DTBW when their four experiments were analysed separately. Instead the real focus of our letter was to contest an apparent 1:1 relationship between the DBM (in grams) and DTBW (in mllilitres) that the authors were able to extract from their data. We have not been able to show this relationship (Nolte et al. 2010a, b; Tam et al. 2009). Instead all our studies show an offset of at least 500 g in this relationship so that a DBM of up to 1,000 g was required before it was possible to detect any DTBW. In our letter, we explained why it is logical to assume that some of the DBM during exercise is from sources other than water and which do not require replacement if the TBW is to be preserved. We are particularly interested in the hypothesis that a fluid reserve of up to 2 l, perhaps existing in the form of unabsorbed fluid in the intestine, may be retrieved as a fluid reserve when the rate of fluid loss from the body exceeds the immediate rate of fluid ingestion. There was substantial interest in this theory during the Second World War (Ladell 1947, 1955, 1965) but the issue was never resolved so that this possibility remains untested. This unresolved questions is not without practical importance. After 1996, the idea took root that athletes should drink to stay ahead of thirst during exercise (Convertino et al. 1996; Armstrong et al. 1996). This was based on the theory that any DBM during exercise is detrimental to both health and performance. Yet, when athletes drink to prevent any mass loss during exercise they usually develop a progressive hyponatremia as clearly shown in two separate studies by Dr Baker and her colleagues (Baker et al. 2005, 2008). We interpret this to mean that some body mass loss is essential during exercise if the serum sodium concentration is to be protected and exercise-associated hyponatremia is to be avoided (Noakes et al. 2005). Indeed changes in body mass alone explain almost all of the variance in the serum sodium concentrations during prolonged exercise (Noakes 2010). That is why it is important to Communicated by Susan Ward.