$Proth$ $theorem$ simply depends on a result which proved by pocklington ;
The result says :
Let $N-1=q^nR$ where $q$ is prime, $n\ge1$ , and $q$ doesn't divide $R$. Assume that there exists an integer $a \gt 1$ such that :
$(1)\ \ a^{N-1} \equiv 1 \pmod N$; and
$(2)\ \ \gcd(a^{(N-1)/ q}-1, N)=1$
Then each prime factor of $N$ is of the form $mq^n+1$ with $m\ge1$
$Proth$ $theorem$ says ;
Let $N=2^nh+1$ with $h$ odd and $2^n \gt h$. Assume that there exists an integer $a\gt 1$ such that $a^{(N-1)/2}\equiv -1\pmod N$ then $N$ is prime
$Proof$ :
$N-1=2^nh$, with $h$ odd, $\,\gcd(2^n , h)=1$ and $a^{N-1} \equiv 1\pmod N$. Since $N$ is odd, then $\gcd(a^{(N-1)/2}-1,N)=1$. By the above result, each prime factor $p$ of $N$ is of the form $p=2^nm+1 \gt 2^n$. But $N=2^nh+1\lt 2^{2n}$, hence $\sqrt N \lt 2^n \lt p$ and so $N$ is prime.
The confusing part of the proof - and the part that I want someone to explain - why does
$$\gcd(a^{(N-1)/2}-1,N)=1\ \ ???????\qquad $$
$Resource$ (the new book of prime number records pages 51 52)
If $\,p \mid (N,\,a^{\large (N-1)/2}\!-\color{#c00}1)\,$ then $\bmod p\!:\ \color{#c00}1\equiv \overbrace{a^{(N-1)/2}\equiv\color{#0a0}{-1}}^{\large\rm by\ \color{#0a0}{hypothesis}}\,$ so $\,p\mid 2,\,$ contra $\,p\mid N\,$ odd.
Update $ $ Per request in comments we elaborate. If the gcd $> 1$ then it has a prime factor $p,$ which is a common factor of $N$ and $a^{\large (N-1)/2}-1.\,$ Therefore, working $\bmod p\,$ we have $$\ a^{\large (N-1)/2}\equiv \color{#c00}{1}$$
The following congruence also holds $\!\bmod p\,$ since by hypothesis it's true $\!\bmod N,\,$ and $\,p\mid N$
$$\,a^{\large (N-1)/2}\equiv \color{#0a0}{-1}$$
Combining the two yields $\, \color{#c00}1\equiv\color{#0a0}{-1}\,$ so $\,2 \equiv 0,\,$ so $\,p\mid 2,\,$ contra $\,p\,$ odd, by $\,p\mid N\,$ odd.
Alternatively $\ p\mid (a^{\large (N-1)/2}+\color{#0a0}{1})-(a^{\large (N-1)/2}-\color{#c00}{1})) = 2.\,$ But one should strive to use congruence methods such as those above (they will prove simpler in more complicated contexts).