Death by Internet

When I read this at the top, I thought to myself, "Wow, I wonder how they showed that! Especially since one would think that isolation would, at least, decrease exposure to flu and cold viruses. I better scroll down and find out."

Then I scrolled down, and:

Other studies have proven that social isolation reduces the effectiveness of tumor-fighting cytokines. Sigman cites a number of studies that have shown that socially active women with breast and ovarian cancer produce more and more-effective tumor-fighters, including the charmingly named Natural Killer cells.

Sigh. Correlation != causation.

It gets worse. Women with fewer social relationships experience strokes at more than twice the rate of those with more social relationships, and those with smaller social networks have narrower arteries - approximately one-third narrower, to be specific.

No, it stays the same. Correlation is still not causation.

Real-world friendships help prevent heart attacks.

They don't even offer justification for this one.

=> Link Bait Headlines

Of course anything can be overdone. Generalizations included.

You really need to look at individual situations. Consider one "Osama bin Laden": I'm betting on (and he's banking on) social isolation hugely extending his life expectancy.