I'm not sure I follow - is this what you're saying:

P1: exposure X is not causally linked to outcome Y if some individuals with lower exposure to X have outcome Y. P2: some individuals with lower exposure to serum cholesterol have heart attacks. C: serum cholesterol is not causally linked to heart attacks.

Because this has some wacky implications:

P1: exposure X is not causally linked to outcome Y if some individuals with lower exposure to X have outcome Y. P2: some individuals with lower exposure to cigarettes have lung cancer. C: smoking is not causally linked to lung cancer.