I'm looking for a lower bound for the falling factorial $$(2n)_{n}:= \frac{(2n)!}{n!}$$
I saw on Wikipedia that $n! > \sqrt{2{\pi}n}(\frac{n}{e})^n$ . So I assume that a possible lower bound could be $$ \frac{\sqrt{4{\pi}n}\left(\frac{2n}{e}\right)^{2n}}{\sqrt{2{\pi}n}(\frac{n}{e})^n}={\sqrt{2}}\left(\frac{2n}{e}\right)^{2n}\left(\frac{n}{e}\right)^{-n}$$
but I am not sure if this is a proper lower bound because the denominator of the LHS depends on $n$ and may be closer to $n!$ than the numerator is to $(2n)!$ making the quotient greater that the true value for $(2n)_n$. I also don't even know if the bound cited in Wikipedia is correct.
Thanks for your help.
We have, through summation by parts:
$$ \log\frac{(2n)!}{n!} = \sum_{k=1}^{n}\log(n+k)= n\log(2n)-\sum_{k=1}^{n-1} k \log\left(1+\frac{1}{n+k}\right) $$ and since $\log(1+z)<z$ for positive $z$ it follows that: $$ \log\frac{(2n)!}{n!} \geq n\log(2n)-\sum_{k=1}^{n-1}\frac{k}{n+k}=n\log(2n)-(n-1)+n\sum_{k=1}^{n-1}\frac{1}{n+k} $$ or: $$ \log\frac{(2n)!}{n!} \geq n\log(2n)-n+1+n(H_{2n-1}-H_{n}). $$
As an alternative, you can use Stirling's inequality: