Table 2.

Numbers of Cases Required to Detect a Significant Association^a in 1, 2, or 3 Fecal Specimens Using an Intraclass Correlation Coefficient From 3 Fecal Sample Populations

Measure and No. of Specimens	Estimated ICCb	Odds Ratioc
		1.5	2.0	2.5	3.0	3.5
Relative Abundance of Phylum-Level Taxa
Actinobacteria	0.44
1		1,496	498	273	190	151
2		1,077	359	196	136	109
3		938	312	171	119	94
Bacteroidetes	0.37
1		1,780	593	325	226	180
2		1,219	406	222	154	123
3		1,032	344	188	131	104
Firmicutes	0.42
1		1,568	522	286	199	158
2		1,113	371	203	141	112
3		961	320	175	122	97
Fusobacteria	0.18
1		3,659	1,219	668	464	370
2		2,158	719	394	274	218
3		1,658	552	302	210	167
Proteobacteria	0.29
1		2,271	757	414	288	229
2		1,464	488	267	186	148
3		1,196	398	218	151	121
Alpha Diversity
PD_tree	0.58
1		1,135	378	207	144	114
2		897	299	163	113	90
3		817	272	149	103	82
Chao1	0.44
1		1,496	498	273	190	151
2		1,077	359	196	136	109
3		938	312	171	119	94
No. of species	0.43
1		1,531	510	279	194	155
2		1,095	365	199	139	110
3		949	316	173	120	96
Shannon index	0.47
1		1,401	467	255	177	141
2		1,030	343	188	130	104
3		906	302	165	115	91
Beta Diversity
Unweighted UniFrac PCoA1	0.81
1		813	271	148	103	82
2		735	245	134	93	74
3		710	236	129	90	71
Weighted UniFrac PCoA1	0.48
1		1,372	457	250	174	138
2		1,015	338	185	128	102
3		896	298	163	113	90

Abbreviations: ICC, intraclass correlation coefficient; PCoA1, first principal coordinates analysis; PD_tree, phylogenetic distance—whole tree.

^a Number of cases (assuming an equal number of controls) required to detect an association at a significance level of 0.05 with 80% power, based on 1, 2, or 3 fecal specimens per subject and the ICC estimated from the fecal samples from the 3 populations. Disease prevalence = 1%.

^b ICCs were the median values of 3 estimates from 3 studies: a National Cancer Institute colorectal cancer study (28, 29), the Human Microbiome Project (30), and a Costa Rica study (Dr. Paula González, INCIENSA Foundation, unpublished data, 2015).

^c Odds ratio for the top 25% microbiome metric versus the bottom 25% microbiome metric.