Columnists

Scientific studies aren’t always perfect

cancer

Two years ago, a study was released showing that exposure to blue light, also known as artificial or LED lighting, increases the risk of breast and prostate cancer, which are the two leading cancers, for men and women, in Nairobi.

A researcher from the study, which was led by the Barcelona Institute of Health, explained that the World Health Organisation’s International Agency for Research on Cancer had classified night work as probably carcinogenic (cancer-causing) to humans. There is also a theory that artificial light disrupts our production of melatonin, which is the chemical that makes us sleep. So, the researchers set off to find out if more exposure to artificial light tied with higher cancer levels.

They concluded that exposure to city lights doubled the risk of prostate cancer and increased the risk of breast cancer by 1.5 times. In the last two weeks, a new study from the institute reported artificial light, and most specifically the light from LED bulbs and from device screens, such as tablets and phones, increased the risk of bowel cancer by 60 percent, too.

All of this data is alarming and does tie with the surge in cancers in our modern world. But the reporting of it also highlights our excessive conclusions from such studies. Not that Barcelona isn’t doing excellent work with these studies — it is. And Nairobi is like many other cities, with the top killer of women in certain age groups being breast cancer, and prostate cancer galloping: we all want to know the cause so we can correct for it.

But dig into the study and it works on correlation alone — and for those who yawned in their statistics classes, that’s the degree to which two things move in the same way. In this case, more city light ties with more cancers, as it happens.

Thus, the scientists mapped how exposed people were to artificial lights based on where they lived and then assessed the prevalence of these cancers.

But the trouble with this correlation is that all those with the most city light also suffer the most vehicle and urban air pollution, more noise, and many other specific city factors. Thus, the cause of the greater cancers in more urban areas could be something else. And correlation studies don’t establish a ‘mechanism’, which is where we see how something disrupts human cells — as in, we can see the way in which a factor affects the human body.

As it is, the reports are surely making me ensure my phone screen is off at night now.

But the whole thing reminds me of a study years back that delivered headlines saying that children who were read stories at night were less likely to get into trouble as teenagers. It was a study I noticed, as I read my kids stories at night, but it kept striking me that the correlation could so easily be picking up something completely different.

Do parents who are in conflict, are alcoholics or drug abusers, grappling with poverty, or a whole range of other parenting-disruptive issues, still tuck their children up into bed each night with a bedtime story? It still seems to me, the bedtime story could be a positive result of other factors that also mean those children will do better as teenagers.

The truth is that any parent who reads stories to their child at night is demonstrating attention, care and respect, valuing their youngster’s wellbeing, and many other things — so the fact that all these can be there at least sometimes may help a lot come that turbulent hormonal and brain development period that unfolds with the teenage years.

But whether our latest correlation is between city lights and prostate cancer, or bedtime stories and teenage mental health, or even between drinking hot tea and throat cancer, the thing about science and studies is that we all need to learn that most of it is about possibilities, not certainties: a ‘maybe’ connection.

Worthwhile, as we continue to understand our world ever better, but not conclusive, yet. For science is a journey. And two things that move the same way only might be because one is causing the other.