Diet Soda Follow-up: Are Diet Sodas Better For You Than Regular Soda?

Artificial sweeteners definitely pre-dated the "obesity epidemic." Saccharin was being used commercially in the early 20th C. and diet sodas were widely available by the 1960s For more on the history of artificial sweeteners, see Carolyn de la Pena's brilliant book _Empty Pleasures_

Soda cans from the 1970s from Found in Mom’s Basement

In response to the recent entry about the association between diet soda and fatness, Jim asked:

Has anyone proved that drinking Diet Soda is better for you than drinking Regular Soda? Does Diet Soda have the same impact on the body as drinking say a glass of water? I haven’t done any research on it and I don’t know if any is out there. I’d really like to see a study of what happens to obese people who stop drinking diet soda and switch to regular.

There’s a ton of research on artificial sweeteners, but I can’t find any studies in which obese people who habitually consume artificially sweetened-drinks were made to switch to sugar-sweetened drinks. That might partially due to ethical/IRB concerns—it’s possible that asking people to consume more sugar than they were previously would be considered a significant health risk. On the other hand, there are studies in which subjects are randomly assigned to consume either artificial or caloric sweeteners, so maybe consuming regular soda falls into the realm of acceptable risk with informed consent.

In those kinds of studies, both “overweight” and “healthy”* individuals who consume regular sweeteners (usually sucrose or high-fructose corn syrup, which are nutritionally equivalent as far as we know) end up eating more calories overall than people who consume “diet,” artificially sweetened foods and drinks. The sugar/hfcs groups also gain weight and fat mass and have negative health indicators like increased blood pressure. I don’t think fatness is bad or that being thin is better, but based on the current available evidence, regular soda appears to be both more likely to make you fat and also worse for your health than diet soda.

*Stupid current labels for BMI categories that don’t correspond at all to actual health outcomes.(1)

A Closer Look at the Studies

This is apparently what Diet Coke looks like in Denmark. Or did in 2009. Pretty! In a 2002 study from Denmark, 41 “overweight” men and women between 21 and 50 years old were assigned to two groups, matched for sex, age, weight, height, BMI, fat mass, fat-free mass and usual amount of physical activity. One group was given sucrose-sweetened dietary supplements (2 g/kg of body weight daily; 70% from drinks and 30% from solid foods) and the other was given artificially-sweetened dietary supplements (an equivalent amount of food and drink by weight sweetened with a combination of aspartame, acesulfame, cyclamate, and saccharin, collectively and individually far below intake levels generally regarded as safe). All the supplements were commercially-available foods and included soft drinks, flavored fruit juices, yogurt, marmalade, and stewed fruits. The researchers note that “great efforts were made before the intervention to find the most palatable artificially sweetened food products on the market for which a matching sucrose-containing product existed.” As some of the artificially-sweetened foods were also fat-reduced, subjects in the sweetener group were given additional butter or corn oil every week.

The study lasted 10 weeks. In addition to the supplements, subjects were free to consume whatever they wanted and as much as they wanted. The subjects visited the lab weekly to pick up the supplements and have urine samples taken (which were used to validate their dietary records). Their height, weight, and fat mass were measured every two weeks. They also kept food diaries that included ratings of their  hunger, fullness, the palatability of the food they ate, and their sense of well-being over the course of each day in the week before the study began, the fifth week, and the tenth week.

Results: The sucrose group ate more calories overall than the sweetener group and got more of their calories from carbohydrates (58% compared to 44%). Both groups decreased how many carbohydrates they were eating in addition to the supplements, but the sugar in the supplements more than made up for the decrease in the sucrose group. The sucrose group gained an average of 3.5 pounds—which was, interestingly, only about half the weight gain that would have been predicted based on how many more calories they were eating. Their activity levels didn’t increase, so the most likely explanation is thermogenesis—i.e. their metabolism changed in response to the increased caloric intake. The group eating artificial sweeteners lost an average of two pounds. In the sucrose group, systolic and diastolic blood pressure increased; in the sweetener group, it decreased. There were no differences in appetite sensations, hunger, or satiety.

Similarly, in a 1990 study done at the Monnell Chemical Sense Center, a group of 30 subjects gained weight during a three-week period when they were given regular soda (sweetened with HFCS) and lost or maintained their weight during the two three-week periods when they were given diet soda (aspartame-sweetened) or no soda. In the regular and diet soda periods of the experiment, they were given 40 oz. of soda to drink every day. In the no soda period, they were told they could consume any beverages as they normally would. They also kept detailed dietary records for the duration of the experiment. The order of the 3-week periods was counterbalanced so some of them got regular soda first, some of them got the artificial stuff first, some of them had no soda for the first three weeks, etc. Here’s what the aggregate changes in their body weight looked like: 

Tordoff and Alleva 1990 in the American Journal of Clinical Nutrition 51: 963-9 (graph appears on 965)

During both the regular and diet soda weeks, they decreased their dietary sugar consumption by an average of 33% (i.e. aside from the sugar in the soda).

Studies like these also point to what I suspect is the more likely explanation why there’s never been a study like the one Jim describes: there’s just not much debate about whether consuming calorically-sweetened drinks leads to weight gain and possible health risks (which shouldn’t be conflated—weight gain is primarily an aesthetic issue, and high levels of sugar consumption may lead to negative health outcomes whether or not they make you fat). What is up for debate is whether artificial sweeteners are a good substitute and likely to promote weight loss or also bad and contributing somehow to weight gain. And if they’re contributing to weight gain, how and how much?

There appear to be three types of theories about why artificial sweeteners might cause weight gain and/or other undesirable outcomes.

Theory #1: Artificial sweeteners might have direct metabolic effects

I don't know what this has to do with anything, I just thought the entry needed more picturesIt’s possible that although they have no caloric value, artificial sweeteners could affect blood sugar or insulin in ways that cause the body to store fat. This is the theory being tested in the study described in the previous entry in which mice consuming aspartame in amounts comparable to an average-sized woman drinking 20 oz of aspartame-sweetened soda per day had higher fasting glucose levels than mice on the same diet minus the aspartame. The effect could be chemical, but seems more likely to be an effect of the sweet taste—i.e. the perception of sweetness might affect the hormones that govern appetite and metabolic rate.

Evidence for this is still extremely scant. Not only is it unclear whether or not the effect is reliable, biologically significant, or occurs in humans; it’s also unclear if it’s specific to aspartame or an effect of all artificial sweeteners, if it scales such that a small amount of aspartame causes a smaller increase in fasting glucose or only occurs at a certain critical level of aspartame consumption, if it only occurs after regular daily consumption over a long period of time or after a single dose, if it affects all people in the same way or only “overweight” people, if it interacts with other dietary conditions (i.e. does it only happen in conjunction with diets high in corn oil, like the ones the mice in the study were fed?), etc.

There are studies involving rats that suggest some kind of metabolic effect of artificial sweeteners might promote weight gain. Rats fed artificially-sweetened yogurt consume more calories than those fed sugar-sweetened yogurt.

However, it seems like it might not work the same in people—or at least that the effect might be smaller. Note that in both of the studies described above, subjects given artificial sweeteners decreased both their overall carbohydrate and dietary sugar intake. Additionally, in a 2001 study done at the Pennington Biomedical Research Center in Baton Rouge, 31 subjects (19 lean, 12 obese) were given sucrose (493 kcal), aspartame (290 kcal), or stevia-sweetened (290 kcal) "preloads" before lunch and dinner on three separate days. Their food intake, satiety (how full they felt), and postprandial (after-meal) glucose and insulin levels were measured. When they had the lower-calorie, artificially-sweetened preloads, they did not compensate by eating more at either the subsequent meals and reported similar levels of satiety as they did on the day they consumed the higher-calorie, sucrose preload.

Theory #2: Artificial sweeteners might have a psychological effect.

Another possibility is that drinking “diet” soda might make people believe that they can afford to eat more or nutritionally worse foods. This is similar to the “health halo” research being done by Brian Wansink and others, which has shown that people are more likely to underestimate the caloric content of foods they perceive as “healthy,” like a turkey sandwich from Subway, than they are with foods they perceive as unhealthy, like a Big Mac. They’re also more likely to order sides with the “healthy” choice that ultimately push the calorie content of the meals higher. Organic and “trans-fat free” labels or even just having calorie counts posted on a menu can have similar effects—triggering people’s dietary conscientiousness seems to cause many people to “treat” themselves to something extra.

However, it’s not clear that the “halo” affect actually influences total or long-term consumption. Thinking they’re getting the “healthier” sandwich may make people more likely to eat a bag of chips at that meal than they would have if they’d eaten a burger, but if that means they’re less likely to have an afternoon snack or they eat less at dinner, it might not affect their weight. I can’t find any studies that measure that.

Theory #3: Artificial sweeteners might change people’s palates

Artificial sweeteners might make people more accustomed to sweetness, which might cause them to eat more sweet things or sweeter things than they would otherwise. Since sweet things and the taste for them are seen as a kind of indulgence and not liking or eating sweet things is often constructed as proof of maturity, masculinity, or self-control, this is often described in morally judgmental terms like “infantilizing our taste sense” or “corrupting the palate.” But it’s not a theory entirely confined to blowhards. In an opinion piece in JAMA published in 2009, David S. Ludwig writes:

Individuals who habitually consume artificial sweeteners may find more satiating but less intensely sweet foods (eg, fruit) less appealing and unsweet foods (eg, vegetables, legumes) less palatable, reducing overall diet quality in ways that might contribute to excessive weight gain.

However, he admits that there’s no research showing this to be true. On the contrary, at least one study has found that people who consume artificial sweeteners regularly are more likely to eat foods generally considered to be healthy and less likely to consume foods generally considered to be fattening. According to a 2006 study done by the American Cancer Society as part of a larger project involving 1-2 million men and women who weigh 40% or more above average for their age and height, those who reported using artificial sweeteners also ate chicken, fish and vegetables significantly more often than non-users and consumed beef, butter, white bread, potatoes, ice cream and chocolate significantly less often. That study also found that artificial sweeteners were associated with weight gain. Given that their diets were apparently “healthier,” the authors conclude: “our weight change results are not explicable by differences in food consumption patterns,” perhaps implying that artificial sweetener might indeed be the culprit.

I think their data suggest something different entirely: people who drink diet soda are more likely to be dieters. They’re eating more of the stuff everyone tells them they ought to be eating to lose weight, and less of the stuff they’re supposed to avoid. It’s not working, and they’re getting fatter anyway, but that doesn’t mean diet soda makes you fat, it could simply mean that dieting doesn’t work.

Not Implausible, Just Not Supported By the Evidence

My suspicion is that if diet soda has any affect on weight, it’s a small one. I think it might be possible that in large amounts (probably 16 oz or more of diet soda per day), some artificial sweeteners might affect the metabolism slightly and lead to people being slightly fatter than they would be if they consumed less or no artificial sweeteners at all. However, I don’t think you’d see the results you see in studies like the ones from Denmark or the Monnell Chemical Sense Center if artificial sweeteners really have a dramatic, immediate effect on weight gain or fat storage.

Of course, that doesn’t mean artificial sweeteners are healthy, just that they probably don’t make you fat. Jury’s still out on the relationships between aspartame and cancer, sucralose (Splenda) and intestinal bacteria, saccharine and neurological function (especially in children), and stevia & its derivatives and DNA mutation. But for what it’s worth, most of the review articles I came across and Ludwig’s JAMA article claimed that concerns about cancer have basically been put to rest.

Of course, there’s still the problem of how they all taste

(1) Broken record footnote: Weight is a poor indicator of health. People in the BMI categories labeled “overweight” and “obese” people are often as healthy or healthier than people in the “healthy” or “normal” BMI category. People in the “overweight” category live longer on average than people in the “normal” or “healthy” category. People who are “overweight” or “obese” who engage in regular physical activity are healthier on basically every measure than sedentary “normal” or “healthy” weight people. The people who are really (statistically) screwed are the “underweight.”