(C) for me

To me, u should try to work on signs of expressions

P + Q > 1/P + 1/Q

<=> P - 1/P > 1/Q - Q

<=> (P^2 - 1)/P > - (Q^2 - 1)/Q

Cases:

o If P > 1, then P^2 - 1 > 0 and P > 0, making (P^2 - 1)/P > 0

o If 0 < P < 1, then P^2 - 1 < 0 and P > 0, making (P^2 - 1)/P < 0

o If -1 < P < 0, then P^2 - 1 < 0 and P < 0, making (P^2 - 1)/P > 0

o If P < -1, then P^2 - 1 > 0 and P < 0, making (P^2 - 1)/P < 0

It's similar for Q...

So what happens with statments?

Stat1
P < Q < 1

We could easily conclude that under 1, P & Q expressions have 3 intervals on which we flip the sign.

INSUFF.

Stat2
P*Q < 1

implies

o P > 0

o Q > 0

but:

o P > 1 and Q < 1

o P < 1 and Q > 1

o P < 1 and Q < 1

All of these possibilities imply different signs for expressions of P & Q.

INSUFF.

Both (1) and (2)
We have

o 0 < P < 1

o 0 < Q < 1

and as we have,

o If 0 < P < 1, then P^2 - 1 < 0 and P > 0, making (P^2 - 1)/P < 0

o If 0 < Q < 1, then Q^2 - 1 < 0 and Q > 0, making -(Q^2 - 1)/Q > 0

we arrive to

(P^2 - 1)/P < -(Q^2 - 1)/Q

SUFF.