The SNP rs17886395, noted as significantly associated with ROP but only under the Dominant model ( p = 0.026) in the complete cohort, is not only marginally significant ( p = 0.052) under this model, but is less significant ( p = 0.093) under the Additive model.
However, when analyzed including covariates GA and duration of oxygen treatment in the analyses, both Additive (p = 0.068) and Dominant (p = 0.059), while not reaching significance, were perhaps marginally significant, as was perhaps alluded to by the shift of all the SFTPA2 SNPs towards marginal significance with the inclusion of these covariates noted earlier.