Multivariate cox proportional Hazards model with Categorical variable

I am working on cox proportional Hazards model recently. I noticed that the variable Sex value was "double" in the dataset of the example of cox proportional Hazards model for censored data you showing us. I believe that Sex should be categorical variable instead of you treat Sex as continuous variable. Is this no difference in Matlab for using categorical or continuous variation? If not, can I treat sex as "f" or "M" in function coxphfit(). My coworker said that there was difference in software statistics for treat Sex as categorical and numerical data to fit Cox proportional hazards function. ps, I use statement exp(stats.beta) to get Hazard ratio, and use exp(stats.beta±1.96*stats.se) to get 95% CI for HR for multivariate model. Is this correct or not?

Any help would be appreciated!

Xiaoying

[Kelly Hennigan]

I'd love to know the answer to this question!