So, does Twitter give you cancer? We've read the study... (updated)

Can Twitter give you cancer? That was the question racing (usually accompanied by the tag "rubbish") over social networks today, based on a new article in the journal Biologist, and then covered by some organisations (sample headline: "How using Facebook could raise your risk of cancer").

The piece in question is by Aric Sigman, a fellow of the Royal Society of Medicine and Associate Fellow of the British Psychological Society, and is entitled "Well connected? The biology of 'social networking'".

It isn't written in the style that I'm familiar with from Nature and Science, the two most eminent peer-reviewed journals. It's rather more like a New Scientist feature, except that those will aim to present both sides of an argument. So this feels (to me, being used to reading science papers for years to try to extract journalistic stuff from them) like science-lite. True, there are lots of graphs and citations. But there's no single experiment that's been carried out. It's more like a meta-study, with a few guesswork-y knobs on.

Sigman's thrust is that "Couples now spend less time in one another's company and more time at work, commuting, or in the same house but in separate rooms using different electronic devices." At this point I'm looking for a citation... though none is forthcoming. But his point feels fair - at least if it said "families" rather than "couples". These days, I'd have thought that couples spend the time in the same room even if they're on different devices.

His next point: ONS data say one-third of adults live alone, "a trend that looks likely to continue." Again, feels right, though no citation.

Then it goes somewhere quite different: into "differential gene expression in lonely individuals", which is the findings of a paper (full text online) from Genome Biology of 2007: "high-lonely" individuals have lower levels of various blood chemicals than those who, um, aren't. Which is an interesting finding in itself. More social contact = reduced morbidity (likeliness to die); less social contact = the other way.

There are plenty of studies showing that real, human physical contact is good for you. Sigman points to a 1998 study that suggested that greater use of the internet "was associated with declines in communication between family members in the house, declines in the size of their social circle, and increases in their levels of depression and loneliness."

OK, that was 1998 though. In fact, Sigman doesn't really have anything to say about social networking systems such as Facebook and Twitter. His article ends with "presiding over a growing body of evidence, we should now explain the true meaning of the term 'social networking'. At a time of economic recession our social capital may ultimately prove to be our most valuable asset."

Er... OK. Nothing about Twitter giving you cancer then? No. There are some older studies which suggest that "women with small social networks [which I don't think means their Facebook friend or Twitter follower count] show more than twice the death rate."

Basically, there's a lot here about social networks - but used in the old-school sense, of the number of people you physically meet, cuddle, interact with face-to-face. What isn't dealt with is the question of whether social networks such as Facebook, MySpace, Twitter, even Skype, which let you interact - at a distance - with people in real time count, for the purposes of sociology, as "real" social networks or as intermediated communication.

And those headlines you've been seeing? All based on a tenuous extension of that Genome Biology study, which Sigman quotes (entirely fairly: he's making a point about living alone). So it's not so much Bad Science as, yes, Bad Journalism that has conflated this into "Twitter gives you cancer".

Except if you go over to the Institute of Biology site, the press release seems to be suggesting that that's roughly what you should think:

Dr Aric Sigman warns us of the dangers of sacrificing old-fashioned social contact for the current trend towards more online interaction. It seems that there is no substitute for face-to-face contact with our family, friends and communities, when it comes to maintaining good health. A Facebook poke cannot replace a good old hug, it seems.

That's not my precise reading after a careful read. You could at least think that it's ambiguous, because it never deals with or cites any papers looking at virtual social networks.

I prefer the observation of one Twitterer the other week about the network's effects: "it's a problem of paucity of language - we don't have a word for friends you haven't met and might never meet in person."

What's needed now is some real research into what happens when you're using online social networks, and to what extent they can supplant face-to-face interaction. Not forgetting that online networks can facilitate real-life meetings, of course.

Update: the Institute of Biology has now made the paper available for you to read.