Buffer under-run or clock sync issues.
There, that was easy to say… much harder to diagnose ;)
To over-simplify, both of those are reasons that the audio has to run to catch up all the time, meaning it occasionally skips a bit. Whenever that generates a rapid zero crossing in the audio waveform (in effect, a big sudden step rather than a sine wave), it will tick.
Buffer underrun happens when your machine is essentially "too busy to do all this at once".
Audio Hijack (& also most DAWs) has a setting to change the buffer size, sliding to the right increases latency by increasing buffer size, to the left decreases.
Clock sync issues tend to happen when one piece of software isn't running at exactly the same clock speed as another, so they constantly slip against each other. Multi-device setups in Audio Midi Setup have a switch to determine which should be the clock master, but I'm not sure I've ever seen anything like that in Audio Hijack. There is the possibility it's too 'consumer' & you might need to look at Loopback instead (or as well) to correct this.
In your DAW, you could probably zoom right into any tick visually & grab a screenshot of what the wave looks like at that exact point (dependant on audio software), which might give additional clues (or might not;).
If your software allows it, you could actually manually draw a new line at that point, making a less sharp transition. Quite often this will remove the tick entirely. It's not perfect, but tbh it can be good enough that no-one will ever spot it.
Things to test, in no particular order
Increase buffer size in Hijack
Try recording after a reboot, with nothing else running except what you need for the recording.
Try recording audio without the screen capture running,
Switch your Scarlett's input frequency in Audio Midi Setup, 44.1 or 48.
Check you have latest hardware drivers for the Scarlett, & the latest macOS sub-version those drivers support. eg, if you're on 10.13, make sure it's 10.13.6
Check you have the latest Hijack.
Test without Hijack, just record into Quicktime.
Test hijack under different file-type outputs, wav vs mp3 etc
Test with mains connected & disconnected on the MBP.
Check without going through the DBX (I doubt this will have any bearing, but you may as well be thorough.)
If I can think of any more, I'll add them.
I had a look at the wave in Cubase, which does show it as quite a complex interference rather than a single squared event - might be a sign that it happened before some other signal processing which 'smeared' the effect.
I managed to just about eliminate it by drawing out the worst of it. I had to go through a few sections to remove the 'smearing'
until I ended up with this…
which is now almost indiscernible.