The study also had another important finding. Most of the sugar we eat is composed of nearly equal parts fructose and glucose, which our bodies process differently. Fructose is primarily metabolised by the liver, and it is often fingered as the culprit in the accumulation of liver fat, whereas glucose generally enters our bloodstream and becomes available for use as fuel. Stanhope found, though, that high-fructose corn syrup, which is 55% fructose and 45% glucose, had effects similar to pure fructose. A meta-analysis also found that excess sugar calories increase liver fat.
But the operative word there might be “excess”.
University of Toronto professor John Sievenpiper told me that, when you add sugar calories on top of normal diets, “that’s when you see all the markers for harm: increase in liver fat, blood pressure, dyslipidemia, all the adverse signals”. What that research doesn’t address is whether all those signals are because the calories are sugar calories or because they’re simply excess calories.
The evidence gets murkier when sugar calories replace other kinds of calories. Stanhope points to research done about 25 years ago that found that, if you give people sugar or wheat starch to the tune of 30% of their calories, the sugar-eaters have higher fasting glucose and insulin. But a similar study found a few differences. And a third study in 1994 found that replacing starch with sugar in breakfast cereal actually improved glycemic responses. Yet another study found that there were no differences between 30g of fructose and 30g of starch.
Overall, it’s hard to find compelling evidence that sugar is much worse than other kinds of carbohydrates. (And I’m not going to open the carb can of worms here: been there, litigated that.)
High-fructose corn syrup is used in many fizzy drinks. Photo / 123rf
Even when we look at soda, which has the reputation of being the worst kind of sugar because it goes down easy and might not trigger satiety, the evidence is mixed. The only study I’ve seen that compares soda to fruit with the same amount of sugar found no changes in weight, blood pressure, or heart disease risk factors in either group. But a study that assigned people to drink a litre of either soda or milk daily for six months found that the soda group, compared to the milk group, had increased liver fat, triglycerides and cholesterol. Unfortunately, studies that compare sugar or soda to protein- or fat-rich foods are few and far between.
Even so, absolutely nobody is defending soda. Sievenpiper (who is upfront about having gotten soda industry funding for some of his work), says that everyone in the public health community agrees that sugar-sweetened beverages (SSBs) are a problem. “They embody calories without any nutritional value.” Even with soda, though, the evidence doesn’t always point to metabolic dysfunction.
The evidence here is neither complete nor compelling. I think it’s likely that sugar, particularly at high doses, can have ill effects. It may raise LDL cholesterol, at least a little. It may contribute to metabolic dysfunction, at least a little. But my rule when the evidence is equivocal is that it probably doesn’t matter very much.
If sugar drives overeating, though, that matters a lot.
There are two main reasons we eat: for hunger and for pleasure. Scientists call those “homeostatic” and “hedonic”.
Stanhope said, “There’s a very prominent hypothesis that our homeostatic system does not recognise the energy from beverages, therefore people automatically gain weight.” She has preliminary results from a study that looks at just that, and the results “show that’s absolutely not true”.
Sugar makes food tastier, which makes it easier to over-indulge. Photo / 123rf
A couple of other studies have looked at this. One, using soda versus cookies, found no differences. The other, using soda versus jelly beans, did find that people ate less of other food when they ate the jelly beans, but not when they drank the soda. But, as Stanhope pointed out, eating 500 calories of jelly beans every day is a tall order. That’s about 125 Jelly Belly beans, which might sound fun on day one, but could get to be a slog by day 28.
Again, if the evidence is equivocal, it probably doesn’t matter very much.
That brings us to what does matter. Remember that part about how we eat for pleasure, not just hunger? Well, sugar simply makes things taste good, and we eat more of foods we like.
A taste for sugar is our evolutionary endowment. Well before they can say, “May I have a cookie,” infants find nonverbal ways – facial expressions, lip smacking – to express a liking for sweetness. And sugar gets even more appealing to us when it’s combined with fat. Our penchant for sweet things conferred a survival advantage because, for nearly all human history from the primordial ooze to maybe 80 years ago, our primary challenge was getting enough calories.
And it still is, in environments where food is scarce. Among the Hadza, hunter-gatherers from Tanzania, the men get 15% of their calories from honey, yet obesity and diabetes are all but unknown to them.
But in an environment of abundance, sugar simply becomes a weapon manufacturers use to get people to eat – and drink – more. Why do you think they put sugar in so many foods? Because it’s their business to find out what flavours or ingredients, and which combinations of them, will drive consumption. (Michael Moss’s book Salt Sugar Fat, is a great read on the subject.)
Manufacturers test levels of sugar, and put the most-liked levels in the food. There’s a body of research, going back decades, devoted to finding exactly how much sweetness people prefer in foods. Because people eat more of foods they like.
Unfortunately, hard evidence for the extent to which sugar drives overeating in a US-style food environment would require us to create a food environment that’s identical except for the amount of sugar in it, and then see how differently people eat when they’re exposed to it. Good luck with that.
Stanhope, whose resume is packed with studies of sugar’s homeostatic effects, believes that “hedonic eating is the main driver of the obesity epidemic, and that our hyperpalatable food environment is the main driver of hedonic eating”. She added: “If sugar promotes overeating, it is via hedonic rather than homeostatic responses.”
Although the definitive study can’t really be done, there’s plenty of research backing the larger idea that it’s temptation, and not hunger, that drives obesity. We eat more when foods are cheap, when they’re right in front of us, and when we’re constantly presented with food cues in the form of pictures, ads and smells.
Sugar, of course, is only one way to make food tempting, and the fact that Americans’ sugar consumption has dropped about 10% in the past quarter-century (to about 91g per person per day), without a corresponding drop in obesity and disease, is evidence that we enjoy eating other things, too.
If you’re getting the sense that the evidence is frustratingly inconclusive, welcome to the world of nutrition research. If the all-over-the-mapness of it surprises you because you’ve read, over and over, about how bad sugar is, that’s probably because so much of the sugar/soda “evidence” comes from observational research, which nearly always shows that people who eat a lot of sugar also have poor health outcomes. If you come here often, you know that I believe that measurement error and confounding prevent that kind of research from drawing causal conclusions, and I rely almost exclusively on controlled trials.
Looking at the entire body of those trials, it’s hard to find evidence that sugar is uniquely bad. But, if you’re trying to eat less, or eat better, or both, and you take inventory of the foods that have the power to undermine your best intentions, it’s a good bet that a lot of them contain sugar.
Which is why nobody, ever, says “Eat more sugar.”