Water activity does not shape the microbiota in the human colon

We recently reported an association between stool consistency as measured by the Bristol Stool Scale (BSS) and major markers of the gut ecosystem, including microbial richness, enterotypes and bacterial growth rates in 53 healthy women. Meanwhile, the link between stool consistency and colon microbiota composition has been confirmed in two large-scale (n>1000) cohorts including individuals of both sexes, different age and varying health statuses. 3 In our original manuscript, we hypothesised on mechanisms that would explain the associations observed. Two potential mechanisms were put forward, namely colon ecosystem differentiation through passage rate variation and reduction of water availability. While the impact of passage rate on ecosystem composition has recently been shown to partially account for BSS-associated microbiota variation, the potential correlation with water availability remains unexplored. Water activity (aw) is a dimensionless variable that reflects the availability of water in a system (aw pure water=1). It differs from water moisture in the sense that the latter quantifies both molecularbound and freely available water. The relation between water activity and moisture content is non-linear and product specific and temperature specific, but has yet to be determined for faeces. Water activity is essential for enzymatic reactions, thus making it a key factor of selective pressure on microbial growth. Its relevance in a driving force for gut ecosystem compositional variation is currently unclear. Here, we measure water activity of 62 fresh faecal samples with different consistencies using a resistive electrolytic hygrometer (LabMaster, Novasina, Lachen, Switzerland; accuracy 0.003 aw; technical variability 0.002 aw), which measures relative air humidity through the resistance of an electrolyte at constant temperature (37°C). We show that water activity in faeces correlates with neither moisture content (Spearman r=0.12, p value >0.05; figure 1A) nor BSS scores (Spearman r=0.03, p value >0.05; figure 1B). Despite large differences in moisture content, which correlate well with BSS scores (Spearman r=0.52, p value <0.001; figure 1C), water activity of fresh faecal samples displays little variation (table 1). The values recorded are consistently higher than 0.97 aw (table 1). Given the relevance of water activity in the context of food spoilage, there is ample availability of data regarding minimal aw values required to allow the growth of specific micro-organisms, From these reports, it is clear that aw above 0.97 provide sufficient unbound water to support the growth of most bacteria. As aw values of fresh faecal samples can be expected not to exceed those in the colon environment, our results indicate that