Quantification of Human Microbiome Stability Over Six Months: Implications for Epidemiologic Studies

Abstract

Temporal variation in microbiome measurements can reduce power. Quantification of this variation is essential for designing chronic disease studies. We analyzed 16S rRNA profiles in paired specimens separated by six months from three studies. We evaluated temporal stability by calculating intraclass correlation coefficients (ICCs). Sample sizes to detect microbiome differences between equal numbers of cases and controls for a nested case-control design were calculated based on estimated ICCs. Across body sites, 12 phylum-level ICCs were greater than 0.5. Similarly, 11 alpha-diversity ICCs were greater than 0.5. Fecal beta diversity estimates had ICCs over 0.5. For a single collection with most microbiome metrics, detecting an odds ratio (OR) of 2.0 would require 300–500 cases when matching one case to one control at P = 0.05. Two or three sequential specimens reduce the number of required subjects by 40%–50% for low-ICC metrics. Relative abundances of major phyla and alpha diversity metrics have low temporal stability. Thus, detecting associations of moderate effect size with these metrics will require large sample sizes. As beta-diversity for feces is reasonably stable over time, smaller sample sizes can detect associations with community composition. Sequential pre-diagnostic specimens from thousands of prospectively ascertained cases are required to detect modest disease associations with particular microbiome metrics.