In a “complete DDS” (includes DAC) spurs are predictable in frequency offset location relative to the output fundamental frequency or FTW. These type spurs include phase truncation spurs, phase-to-amplitude spurs, DAC harmonics, and DAC harmonic fold back products. One way to determine if a spur is DDS related is to simply change the FTW. If the spur in question remains fixed in frequency offset for both old and new FTWs, the spur is probably not DDS related, less one exceptions below.
Most DDSs have an internal REF CLK multiplier (PLL) option. If enabled, reference clock spurious would be fixed in frequency offset to all FTW changes. The is same true for any stand alone integer N PLL synthesizer.
For the easiest test, pick FTWs that are exact integer to the system clock like 1/4th and 1/8th FTWs. With integer FTWs settings, all DDS spurs are harmonically related to the fundamental frequency. So, there should be no significant spur energy present other than harmonics of the fundamental. If so, the spur is not DDS related with the exception above. In addition, the integer settings above produce no phase truncation spurs.
If you choose to use non-integers FTWs for the test, keep the FTWs in the middle (15% to 25%) of the DDS range relative to the system clock. That should avoid fold back issues of the spur around DC and the nyquist limits if the fundamental is too close to either limit.
After a FTW change, if the spur remains fixed in frequency offset to fundamental change, likely sources are spurs on the Reference Clock source with and without the internal PLL enabled, and/or spurs on the power supplies or some other source in the system. AN-927 from ADI has pictures based on reference clock spurs and power supply related spurs.