Table 7.

Mediation analysis to test for the indirect effects of treatment type on falls via VEST scores from the SOT.

	b	SE	p-value	Odds Ratio*	95% CI Odd Ratio
Test of indirect path from adjuvant endocrine therapy only† as the independent variable
Adjuvant endocrine therapy only to VEST score	−6.43	6.84	.35	n/a	n/a
Direct effects of VEST score on falls	−.04	.01	.02	.96	[.94,.98]
Total effect of adjuvant endocrine therapy only on falls	.47	.61	.45	1.60	[.48,5.29]
Direct effect of adjuvant endocrine therapy only on falls	.31	.65	.63	1.36	[.38,4.87]
Partial effect of chemotherapy only on falls‡	−.23	.78	.77	.79	[.17,3.66]
Indirect effects of adjuvant endocrine therapy only on falls through VEST score	.22	.30	n/a§	1.25	[.76,2.64]||
Test of indirect path from chemotherapy only‡ as the independent variable
Chemotherapy only to VEST score	−19.41	7.72	.02	n/a	n/a
Direct effects of VEST score on falls	−.04	.01	.02	.96	[.94,.98]
Total effect of chemotherapy only on falls	.40	.69	.57	1.49	[.39,5.77]
Direct effect of chemotherapy only on falls	−.23	.78	.77	.79	[.17,3.66]
Partial effect of adjuvant endocrine therapy only on falls†	.31	.65	.63	1.36	[.38,4.87]
Indirect effects of chemotherapy only on falls through VEST score	.68	.49	n/a§	1.96	[1.05,6.38]||

CI: confidence interval

^*Odds ratios reported only for logistic regression analyses

^†Comparison of participants who only had adjuvant endocrine therapy to participants who had chemotherapy + adjuvant endocrine therapy

^‡Comparison of participants who had chemotherapy only compared to participants who had chemotherapy + adjuvant endocrine therapy

^§Normal theory tests are not available for models with binary outcomes

^||Confidence intervals are bias corrected and accelerated