I am getting different number of dysregulated genes when performing lfcSrink() on the DESeq run. Sample size: I have 2 groups High(n=16) and Low(n=8). Following is my code for creating the DESeq object:

# Creating DESeq dataset
dds_risk <- DESeqDataSetFromMatrix(
  countData = count.mat %>% 
    column_to_rownames("Geneid"),
  colData = sample.pdata %>% column_to_rownames("Sample.Name"),
  design = ~ Risk.group
)

# Setting reference
dds_risk$Risk.group <- relevel(dds_risk$Risk.group, ref = "Low")

# Run DESeq (quiet)
dds_risk_run <- DESeq(dds_risk)

Now, when I am performing the lfcShrink() using following code I get 159 upregulated and 265 downregulated genes

High_low.result <- results(dds_risk_run, alpha = 0.05, lfcThreshold = 0.585,
                           contrast = c("Risk.group", "High", "Low"))
High_low.resultShrink <- lfcShrink(dds_risk_run,
                                   coef = "Risk.group_High_vs_Low", 
                                   type = "apeglm")

But when I use the following code I am getting only 8 upregulated genes.

High_low.result <- results(dds_risk_run, alpha = 0.05, lfcThreshold = 0.585,
                           contrast = c("Risk.group", "High", "Low"))
High_low.resultShrink <- lfcShrink(dds_risk_run,
                                   coef = "Risk.group_High_vs_Low", 
                                   type = "apeglm",
                                   res = High_low.result)

The only difference in those steps are that I am supplying the results along with DESeq run object and coef in the later method. Which as per documentation should give same result when the results are not supplied, but in my case I am getting different results. Can anybody explain me why this is happening?