Abstract Subdwarf A-type stars (sdAs) are objects that have hydrogen-rich spectra with surface gravity similar to that of hot subdwarf stars but effective temperature below the zero-age horizontal branch. They are considered to be metal-poor main-sequence (MS) stars or extremely low-mass white dwarfs (ELM WDs). In this work, using the stellar evolution code Modules for Experiments in Stellar Astrophysics, we investigate the sdAs formed both by the evolution of (pre-)ELM WDs in double-degenerate systems and metal-poor MS stars with single evolution models. We find that both of the evolutionary tracks of ELM WDs and metal-poor MS stars can explain the observation properties of sdAs. However, the proportions between these two populations are uncertain. In this work, we adopt the method of binary population synthesis of both ELM WDs in the disk and metal-poor MS stars in the halo to obtain their populations at different stellar population ages and calculate their proportions. We find that the proportion of metal-poor MS stars to sdAs for a stellar population of 10 Gyr is ～98.5%, which is consistent with the conclusion that most sdAs (>95%) are metal-poor MS stars. And the proportion of ELM WDs (metal-poor MS stars) to sdAs increases (decreases) from 0.1% (99.9%) to 20% (80%) with stellar population ages from 5 to 13.7 Gyr.