Question: R: How to test for trend of risk scores

1

magda.meier •

**20**wrote:I have some case/control data like so (many more rows)

```
ID score phenotype
1 0.6482772 0
2 0.9519315 1
3 2.7689036 0
4 1.9974466 1
5 1.6287841 1
6 1.9993825 0
```

I performed a logistic regression on pheno~score to determine if an increase in scores leads to increased odds ratio (a measure of disease risk)

I now want to do a trend test in R to test whether there is trend. This trend test would just give me one odds ratio and a p-value. Does anyone know which R package or function I could use? I've looked around and I'm struggling to find a simple package that can help me do this.

Thanks

ADD COMMENT
• link
•
modified 13 days ago
by
Biostar ♦♦

**20**• written 17 days ago by magda.meier •**20**
As phenotype is either 0 or 1 could you just do a fisher test as :

Could also be nice to plot the data as boxplots (x is the data.frame containing your data):

5.3k