It is a follow-up question linked to this post: C: Could the batch effect be a set of numeric values in DEseq2?

The numeric covariate here is the load and it is likely to have a linear effect:

X,Time,Treatment,load
sampe1,Treat1,1
sample2,Treat1,7
sample3,Treat1,23
sampe4,Treat2,4
sample5,Treat2,59
sample6,Treat2,10

I built a LRT to test to find genes related to load that is affecting gene expression:

full=~load + Treatment
reduced=~Treatment

DEload <- results(model, name="load", test="Wald", alpha=0.05)
summary(DEload)

I got a list of DEGs related to load looks like this:

       baseMean log2FoldChange      lfcSE        stat    pvalue         padj
      <numeric>      <numeric>  <numeric>   <numeric>    <numeric>    <numeric>
 gene1   1.301526     0.12255733 0.05282894   2.3198898 2.034684e-02 0.0746050758
 gene2   5.850844     0.11564624 0.02771495   4.1727025 3.010078e-05 0.0004198718
....

I understand that in some comparisons when the model is testing differential expression between two factor conditions the log2foldchange showes the foldchange of the gene expression in one treatment compared to another. But how do I interprete the log2foldchange in here since I am testing a numeric covariate?

Any help can be useful!