OK just tried holding the nano in reset by setting it to ground same issue.
So, just to confirm, that was with the Nano properly powered and Nano ground connected to target ground?
If not, please ignore the following, make it so and re-test.
I''ve just double-checked the ATmega328P datasheet to be certain I'm not misleading you:
15.1. Resetting the AVR
...
The I/O ports of the AVR are immediately reset to their initial state when a reset source goes active. This does not require any clock source to be running.
...
As all DDRx and PORTx registers are zeroed as their initial state, the I/O pins you are using for SPI *should* be tri-stated if the Nano is held in reset. If /SS (your CS) isn't then either there's something wrong with your circuit or with that pin of the Nano. Check that pin is floating with the Nano held in reset and no target connected (monitor the pin voltage with a DMM and see if it can be pulled both high and low with a 10K resistor alternately to Vcc and Gnd).
There is a small possibility its an AC loading effect caused by the extra capacitance of your programming cable and programmer circuit, that appears when the target accesses the memory becauuse its either got less drive or operates at a higher transfer rate (or both) than your programmer, so is more sensitive to load capacitance, however I would expect such trouble to be associated with the SPI clock or data lines not /SS, as they have far more frequent timing critical transitions than /SS.