Statistical Design
The calculation assumed 12-week resolution rates of 80.7% for rf-PDT and 37.9% for placebo. The study is powered to detect associations between the OGTT z-score composite and 1-month human milk LA and DHA concentrations with 80% power. The…
2 sources - 11 claims
The calculation assumed 12-week resolution rates of 80.7% for rf-PDT and 37.9% for placebo. The study is powered to detect associations between the OGTT z-score composite and 1-month human milk LA and DHA concentrations with 80% power. The previously treated cohort is exploratory and not powered for formal hypothesis testing. The sample size calculation used Chan et al. 2008 acute CSC data as the best available placebo-controlled evidence. Primary analyses use cross-sectional multivariable linear regression models. Secondary analyses use false discovery rate correction for multiple comparisons.