It's a triangle. So, they are related by the simple equation ∠DHA+∠HAD+∠ADH=180o. You may replace one with the other.

You may interpret a hydrogen bond as a coordination bond, i.e. a bond with a polar and a certain covalent contribution, somehow similar to semipolar bond, but with a significantly lower covalent contribution. If you look at the respective molecular (LCAO) orbitals you may observe contributions from donor, H, and acceptor, qualifying the hydrogen bond as a 3-center bond with the hydrogen nucleus modifying the donor-acceptor overlap. In a formal MO representation you have an occupied nonbonding orbital (lone-pair) of the acceptor interacting with an unoccupied and anti-bonding H-D orbital of the hydrogen bond donor.

Besides the spatial distance of donor and acceptor, the relative orientation of the nonbonding hydrogen bond acceptor orbital and the anti-bonding hydrogen bond donor orbital are important for the overlap, the electron density in the donor-acceptor region, and the strength of the resulting hydrogen bond. The actual "average" position of the hydrogen nucleus results from this overlap. An extended H-D bond length is a clear indication of an existing hydrogen bond.

As for the angle criterion, it is often helpful to look at the acceptor's precursor to identify the spatial orientation of the occupied nonbonding orbital of the acceptor. In the case of a carbonyl group, e.g., this would be the carbon atom and the C=O---donor angle should be as close as possible to the spatial orientation of the oxygen's lone-pair(s) for "good" hydrogen bonding.

I recommend looking at other papers that list hydrogen bond angles in the journal you wish to publish in and go with what seems to be the norm for that journal. I personally always measure from heavy atom to heavy atom with the Hydrogen in the middle, so D-H----A (of course A----H-D would/should give you the same result so you could probably go with that as well). But the goal is 180 degrees.