Gut microbiota ecology: Biodiversity estimated from hybrid neutral-niche model increases with health status and aging

and consistent results are obtained with the Hill1 and Hill2 numbers (Mann-Whitney U-test p- value = 0.043 and 0.047, respectively). Overall, the outcome of the pairwise comparisons between the three groups of the Bia- gi&Schorr data set suggests that the estimate of biodiversity derived from the 2NB model is consistent with classical indices, while having a higher statistical power (Fig 4 and S8 Fig). When comparing Healthy Italian and Healthy Hadza, in fact, the test based on θrare achieves the lowest p-value, followed by θabundant; when comparing Healthy Hadza and DS Italian, the smallest p-value is obtained with θrare, followed by Shannon and by Hill1 and θabundant, tied for third place; and when comparing Healthy Italian and DS Italian subjects, the smallest p-value is obtained with θrare, followed by Hill1 and Hill2, while none of the other indices detects a sta- tistical significant difference. Such increase in statistical power allows us to obtain an interseting result that is not obvious when using classical statistical indices: DS subjects have a lower GM biodiversity than age- matched controls. Notice that DS is associated with accelerated aging [33] and that the global trend of the splines regression model reported in the previous section indicates a general increase of GM biodiversity with age. This suggests that the increase of GM biodiversity with age that we observed in such model is related to healthy aging and that a decrease of GM biodi- versity is associated with unhealthy phenotypes rather than with a slower aging.

This interpretation was corroborated by the investigation of the GM biodiversity of the healthy and unhealthy elderly subjects from the ELDERMET data set at T0 (see the Materials and Methods section for details on the classification of subjects). Fig 5 shows that while healthy Fig 2. Spline model for θrare. The splines model corresponding to each individual data set is shown. In each plot, dots correspond to subjects, the x-axis is given by the subject age and the y-axis is the standardized value of θrare. Black lines and gray shadows represent the data set specific regression lines and confidence intervals. https://doi.org/10.1371/journal.pone.0237207.g002 PLOS ONE Gut microbiota biodiversity and healthy aging PLOS ONE | https://doi.org/10.1371/journal.pone.0237207 October 30, 2020 8 / 23