I am wondering whether I can do a tobit regression as linear regression for a continuous dependent variable where some measurements are below a detection limit. The distribution of my data is currently very right skewed because of a portion of the datapoints being <5. I want to regress several variables such as age on this dependent variable. I see that tobit is often used in survival analysis, but in my case I actually have cross-sectional data. Is tobit regression a correct method here?

I first simply removed those below detection values and log transformed my data which made it somewhat gaussian, but now that I think it through, I would like to look at the <5 values too.