Table 2.

Analyses of Associations of Body Mass Index With Beta-Diversity at Taxonomic Level 6 (Genus)^a

Dissimilarity	Standard Reference Method							MiRKAT
						2-df   P Value	5-df   P Value	P Value
Bray-Curtis
Mainb	2.27(0.98)c	7.96(5.12)				0.0406c
Quadraticd	1.62(1.14)	4.15(10.67)	6.84(5.89)	−38.98(73.08)	−16.06(47.87)		0.0012c	0.0102c
Unweighted UniFrac
Mainb	2.67(3.37)	−1.43(3.88)				0.6696
Quadraticd	−1.41(3.90)	0.83(4.12)	62.72(41.06)	132.7(85.71)	−41.99(94.86)		0.0116c	0.0020c
Weighted UniFrac
Mainb	2.89(2.25)	1.90(6.49)				0.1924
Quadraticd	−3.19(3.34)	−13.41(9.53)	12.55(20.93)	−223.3(113.60)c	−131.6(102.90)		2.03 ×10–5c	0.0347c

Abbreviations: df, degrees of freedom; MiRKAT, microbiome regression-based kernel association test; se, standard error; UniFrac, unique fraction.

^a Analysis of data from the American Gut Project (12). Regression coefficient estimates and their standard errors (se) are denoted, for example, by . All analyses include age and sex in the models to adjust for potential confounding by these factors.

^b Main effects for distance to stool and distance to nasal samples (X_stool and X_nasal; see Methods).

^c Statistically significant association at 0.05 level.

^d Model also includes (X_stool)², (X_nasal)² , and (X_stool) (X_nasal).