You need to use logistic regression here, since the dependant variable is categorical.

Now if you have dichotomous dependent variable, use binary logistic regression and if you have multiple dependant variable, use multiple logistic regression.

Slight adjustment to Ms. Chowdhuri's reply. First where she says multiple logistic regression she means multinomial logistic regression. Second if the DV is ordinal you want ordinal logistic regression. Useful set of notes is attached. Best, David Booth

PS if there is more than 1 DV in the regression equation then you have multivariate multiple logistic Regression. Google this term for details.

Shady

Use GLM with binomial familiy and log link to get lnRR. Logistic regression gives you lnOR. By design (cohorte study) RR is the best choice

Diego Fernando Rojas Gualdrón thanks for answer, what do you mean by GLM and RR?

Hi Diego, thank you for your response. Would you like to explain GLM and RR?

Shady Abohashem and Praheli Dhar Chowdhuri

GLM stands for Generalized Linear Model. It is an expansion of the general model where you can set the distribution of the dependant variable and the link function with independent variable(s).

For example:

linear regession would be GLM with normal distribution and identity link function.

Logistic regression would be GLM with binomial distribution and logit link function

logistic regression gives you lnOR as regression coefficients (if you exponentiate them you get OR Odds Ratios). OR should be used with case control studies. Yes I know that people use them even with RCT.

If you want to get Risk Ratios RR, the appropiate measure for cohort studies, then use GLM with binomial family and log link function. It gives you regression coefficients expressed as lnRR

Check SPSS overview for GLM and look in its in-software help.

https://www.ibm.com/support/knowledgecenter/en/SSLVMB_23.0.0/spss/advanced/idh_idd_genlin_typeofmodel.html