Protons bound to heteroatoms are so-called exchangeable protons, and can get replaced by deuterium from your solvent.
Due to the rapid exchange, they can also be a part of the water (or other OH signals) if your solvent contains a lot of water.
Furthermore, they can be broadened enough to disappear into the baseline due to chemical exchange.
When the water - OH exchange rate is similar to the peak separation (in Hz) of the water and the OH peak location, the signals become very broad. This means it is possible for protons involved in chemical exchange to appear differently in spectra measured at different B0 strengths.
Exchange protons such as the OH of a carboxylic acid are sometimes missing from the proton NMR spectrum because they are exchanged on the time scale of the NMR experiment
Anything said above is perfect. Just to mention the most trivial case for completeness: if your solvent is D2O many exchangeable hydrogen atoms of most analytes will disappear within a very short time. They are replaced by a "1H NMR invisible D".
Alfred
After the addition, the proton of the hydroxyl group will undergo a swift exchange with a deuterium atom. Since deuterium atoms do not generate peaks in a standard NMR spectrum, the initial peak corresponding to the hydroxyl group will vanish.