Table 4.

Association of SNPs in APC with ASE outcome^*

	Cases and controls combined		Controls only		Cases only
SNP	Anova p-value	Bartlett’s test p-value	Anova p-value	Bartlett’s test p-value	Anova p-value	Bartlett’s test p-value
rs41115	0.0091	0.066	0.5907^	0.803	0.0058	0.131
rs41115 (dominant)**	0.0268#	0.015	0.9088^	0.802	0.0013	0.116
rs459552	0.8310#	0.013	0.9640^	0.653	0.7220#	0.025
rs467033	0.1060^	0.212	0.1752^	0.989	0.2535	0.289
rs467033 (dominant)**	0.0342^	0.222	0.1752^	0.989	0.0973	0.302
rs971517	0.0010	0.084	0.1641^	0.935	0.0103	0.148
rs2431238	0.7139^	0.097	0.9929	0.671	0.2476#	0.041
rs2431238 (dominant)**	0.4118#^	0.019	0.9049	0.470	0.2476#	0.041
rs2431507	0.0168^	0.190	0.0920^	0.887	0.0973	0.302
rs10428710	0.8598^	0.063	0.9337^	0.840	0.5819#	0.013
rs10428710 (dominant)**	0.7851#	0.019	0.9898^	0.775	0.5819#	0.013
rs13358140	0.1509^	0.178	0.1336^	0.591	0.6526	0.066

^*Analyses were done on untransformed data using either ANOVA, where appropriate, or Kruskal-Wallis (as noted). This was because transformation failed to normalize data that deviated from normality.

^**In situations where there were few homozygotes of one class, analyses were performed using all three genotypes, as well as on heterozygotes pooled with the rare homozygotes (dominant model).

^#Kruskal-Wallis test used because of deviations from equal variance

^{^}Kruskal-Wallis used because of deviation from normality