You can increase your chances of living forever by avoiding french fries, according to the Telegraph. Eating chips ‘doubles your chances of death‘, according to the newspaper. Step away from the deep fat fryer.
Epidemiological studies often take groups of people of a certain age and observe how many of them die over a certain number of years. This can be useful in identifying the causes of premature mortality, but it does not tell us anything about our ‘chances of death’ which remain 100 per cent despite the best efforts of the health lobby.
Should we avoid chips on the basis of this study? I wouldn’t be too hasty. Published in the American Journal of Clinical Nutrition, it is a fairly average study by the standards of nutritional epidemiology. It does not have any glaring flaws that set it apart from the others, but that is not high praise. Most epidemiological studies are useless and nutritional epidemiology, in particular, is a cesspool of contradictory findings and confounding factors.
This is not the fault of the researchers. Sending people a food survey once in a blue moon and then waiting for them to die is a weak methodology at the best of times, but what else can we do? One thing we could do is stop funding this junk science, but then newspapers would have nothing to print next to the cartoons.
So what does the study say? The researchers set out to find whether potato consumption increases the risk of premature mortality. The presumption is that it doesn’t because, as they note, a systematic review published last year did not find any ‘convincing evidence’ to link potato consumption with heart disease, diabetes or obesity.
The researchers used this website to find 4,000 people between the ages of 45 and 79. They only selected people who had knee osteoarthritis or were at high risk of developing knee osteoarthritis. It is not obvious why they focused on this disease, but obesity is a major risk factor for knee osteoarthritis so it is likely that the group contained a higher proportion of obese people than the national average.
These 4,000 people were sent a survey asking them about their diet, including questions about how often they ate fried or unfried potatoes. As you might expect, most of them ate potatoes at least once a week.
The researchers then waited to see how many of them died. After eight years, 236 of them – one in twenty – had joined the choir invisible. For some reason, the researchers were unable to ascertain what they had died of, but they nevertheless set about analysing the data.
It turned out that the death rate among those who ate potatoes three or more times a week was three times higher than among those who rarely or never ate potatoes. However, this was not the only difference between them. The potato enthusiasts also tended to be older, male, less well educated and less healthy. The apparent association between potato consumption and premature mortality disappeared when the researchers adjusted their model for age, sex, race, body weight, income, physical activity, alcohol consumption and other factors.
But when the researchers looked specifically at fried potato consumption, adjusting the model had less of an effect – the association fell from a threefold increase in risk to a twofold increase in risk. Or, if you prefer, it ‘doubled the chances of death’.
This might sound alarming but, as Nick Ross used to say, don’t have nightmares. As the authors note, ‘consumption of potatoes was very high’ among the 4,000 people in this study and yet 95 per cent of them were still alive by the time it was concluded. The authors also note that the kind of people who eat a lot of fried potatoes might also be the kind of people who eat a lot of ‘red meat, salty foods, and sugar-sweetened beverages, which may increase the risk of death’. This seems to me more than a mere possibility. It is highly likely, and yet the study was unable to control for it.
It is also likely that the people who never ate potatoes had different characteristics from those who regularly ate french fries. I would never be so rude as to describe people who abstain from potatoes as being odd, but they are clearly not average. The researchers did their best, given the limited information available to them, to control for such variables but, in the final analysis, the seriousness with which you take these findings depends on how confident you are that they have made the right adjustments to their model.