There aren't any natural number $n$ for which the given condition is satisfied. Here is how I proved it:
For $\sqrt{n-1}+\sqrt{n+1}$ to be rational, $\sqrt{n-1}$ and $\sqrt{n+1}$ must individually be rational as the sum of two positive numbers can't be rational if at least one of them is irrational.
For $\sqrt{n-1}$ and $\sqrt{n+1}$ to be individually rational, $(n-1)$ and $(n+1)$ must be perfect squares. We can see that the two numbers differ by $2$: $(n+1)-(n-1)=2$. This is not possible as consecutive squares differ by at least $3$. $\blacksquare$
The book I am using proves the same conditions in a more involved way:
If $\sqrt{n-1}+\sqrt{n+1}$ is rational, $\frac{1}{\sqrt{n-1}+\sqrt{n+1}}$ must also be. Multiplying by $\frac{\sqrt{n-1}-\sqrt{n+1}}{\sqrt{n-1}-\sqrt{n+1}}$, we get $\frac{\sqrt{n-1}-\sqrt{n+1}}{2}$ is rational $\implies \sqrt{n-1}-\sqrt{n+1}$ is rational.
The proof then proceeds to do what I did but for both $\sqrt{n-1}-\sqrt{n+1}$ and $\sqrt{n-1}+\sqrt{n+1}$.
There is also an answer on Mathematics that proves the same in a way similar to the book prove I just mentioned.
Is the way I have proved the same incomplete or wrong?
My question is not duplicate because it is asking why my proof is wrong. The linked duplicate asks for a proof; I have provided the proof, both mine and the one found in book, with the question.
The sum of two irrational numbers can be rational (e.g. $x,\,q-x$ with $q\in\Bbb Q\not\owns x$). Their strategy was to note that, since $\sqrt{n-1}\pm\sqrt{n+1}\in\Bbb Q$, taking linear combinations thereof gives $\sqrt{n\pm 1}\in\Bbb Q$.