Here is a UV/Vis spectrum of trehalose. It appears that it has a max at 626 nm which is in the visible range. Since the compound does not have a chromophore (double bonds) I would doubt that is absorbs in the UV.
It's probably easier to use another detector. If you have an evaporative light scattering detector or a chared aerosol detector then you can try that. If your HPLC method does not use a gradient then refractive index detection will work. If you have access to an instrument with a pulsed amperometric detector that will probably be the most sensitive option without using any kind of derivatising agent. Of course there's always the MS option if you have access to one of those.
Otherwise you might want to change technique completely and consider GC-FID after derivatisation....