The following originally appeared on The Upshot (copyright 2015, The New York Times Company).
Readers responded in great number to my recent column on artificial sweeteners. Many comments were on the same topics, which I’ll address here.
1. What about the “Nature” study that showed that artificial sweeteners alters the gut microbiome and leads to diabetes?
This is the most important comment to address. Many readers said that I ignored a crucial study published last year in a big journal that got a lot of attention. I didn’t ignore it. But let me explain why I left it out.
The manuscript described many experiments. First, researchers showed that mice fed aspartame, sucralose or saccharin had higher blood-glucose levels than mice who were fed just water or sugar. Second, they showed that if they sterilized the guts of mice, or cleared out their “microbiomes,” and then did bacterial transplants from mice who were fed artificial sweeteners, those mice with the new bacteria had higher blood sugar levelsthan those who got transplants from mice fed sugar. Third, they showed that people who use artificial sweeteners have different levels of Enterobacteriaceae, Deltaproteobacteria and Actinobacteria in their guts than those who don’t.
The final study involved seven people, five men and two women. They were healthy people who didn’t use artificial sweeteners. They were given the Food and Drug Administration’s maximum allowable amount of saccharin for six days. Four of the seven developed “abnormal glucose responses,” and three didn’t. The four people who developed this also showed changes in the bacteria in their guts.
So what we have are two mice studies, one relatively small cross-sectional analysis of people that can’t establish causality and didn’t appear to control for anything other than B.M.I., and a final prospective study of seven people followed for a week. I think that the results of this manuscript are worthy of follow-up and further research, but they don’t merit inclusion as to what we can say artificial sweeteners do to humans in practice. With so few participants, it’s hard to discuss validity at all.
There are also a lot of holes here. As reported in The New York Times when the study came out, no one knows how the artificial sweeteners affect the bacteria in the gut. No one knows why the three different artificial sweeteners resulted in similar changes when they are completely different molecules. And no one really knows how the interplay between bacteria in the gut, or the “microbiome,” affects health. There are some interesting theories, but they are nowhere near conclusive.
2. What about other studies you ignored?
I try to cite meta-analyses and systematic reviews in my columns. I like to talk about the collected body of research instead of trying to pick a study. If I cite individual studies, I try and cite randomized controlled trials or large cohort studies. If I do address individual studies, I still try and place them in the context of the larger literature.
I also favor studies of humans over rats, and real outcomes over process measures. So if you’re bringing up a rat study that you found, it’s likely I didn’t spend a lot of time on it. This is especially true with artificial sweeteners, which are ubiquitous and easy to study in people. If you bring up a study of a process measure (like some temporary enzyme level), it’s likely I considered it less than a randomized controlled trial of weight gain or loss.
I make every effort not to cherry-pick. I’m not perfect, and I imagine that at some point I will miss a crucial study somewhere. I don’t think that’s the case this time.
3. How can a pediatrician let his kids drink diet soda?!?!
I didn’t anticipate the amount of anger that my last paragraph in the column would create. I don’t think that letting my children drink diet soda once in a while makes me a monster, but apparently some of you think it does.
My children’s food intake basically follows the rules for healthy eating that I described in a previous Upshot column. They eat a lot of vegetables. They eat a lot of unprocessed foods. They drink a lot of water. Once in a while, I let them have a diet soda if they like. Once in a while, I let them have dessert, too. Moderation seems to work, and they’re all doing great.
4. Why does anyone need artificial sweeteners, or diet soda, at all?
They don’t, obviously. They also don’t need beer, Scotch, cheesesteaks, pizza or apple pie. But those are five things that I love, and I like to have them once in a while. If you choose to forgo every single thing that isn’t perfectly efficient and nutritious, then that’s your choice. Most people aren’t going to do that. I’m not, at least.
Some people drink nothing but water. More power to them. I’m not convinced that allowing yourself the occasional delicious treat is really that unhealthy for most people. The key is to keep it to a minimum. If someone enjoys the occasional artificially sweetened drink, then I don’t see solid evidence that that’s a problem. But some people overdo it, and that leads me to the next question.
5. Some people drink a lot of diet soda. Isn’t that bad for them?
Maybe. But we have to compare it to the alternative. I see a lot of children in clinic who drink gallons of juice, soda or milk, and who have a weight problem. The first thing I try to do is to eliminate the empty calories of sugar-sweetened beverages. Getting them to go to water is often very hard. I know, some of the commenters did it without any difficulty at all, but I don’t think that’s the experience of lots of people.
If I can get a child to switch to sugar-free lemonade, at least as a start, I’m eliminating hundreds of calories a day of added sugars. It’s a comparative thing. I’m not saying artificial sweeteners are safer than nothing. I’m saying that if I had to choose between the diet drink and the one with added sugars, I’d go with the diet drink. But I agree that we should limit the consumption of all of these things to a minimum, which would be consistent with my previous recommendations on eating.
