Yogurt, Cancer, Scientific Bias

"Whatever belief or suspicion I have expressed in the past, I'm willing to subject them all to the scrutiny of unbiased science." (RFK Jr., 2/18/2025, speech to HHS employees.)

Noble words. But what counts as "unbiased science"?

Is it, for instance, decades of peer-reviewed studies showing no link between vaccines and autism?

Or is it the study RFK Jr. mentioned during his confirmation hearings that does propose a link?

The "study" RFK Jr. cited is actually a blog post. The lead author and editor have a history of scientific malfeasance and retracted papers; the "study" itself is hopelessly flawed. (I discuss the details here.)

RFK Jr. seems to be one of those people who defines "unbiased science" as "science that reflects my biases."

My own liberal biases are on display here. I believe that RFK Jr. is unqualified to be HHS secretary, and that his confirmation poses a grave threat to American public health.

However, scientific bias isn't inherently partisan. Even the crunchiest liberal scientists may let their ideology lead the data rather than the other way around. In other cases, scientific bias doesn't seem political in the first place.

This newsletter illustrates the point. I'll be discussing a new study from Harvard researchers suggesting that eating yogurt prevents a certain type of colon cancer.

Although the study is biased, I'm not sure I'd blame liberal ideology. Call it a pro-yogurt bias. A very strong one.

All the same, I have no doubts that yogurt is healthy in many respects. After discussing the study, I'll conclude with some favorable remarks on yogurt and other fermented foods.

Some context

The authors of the new study don't make the broad claim that yogurt prevents cancer (although, judging from an email to me, the lead author appears to believe this). Rather, the claim is that a particular ingredient in yogurt reduces the risk of a particular type of cancer. I want to share a little background on each.

Bifidobacterium

Part of what makes yogurt creamy and delicious is the bacteria used to ferment the milk. These are the "live and active cultures" sometimes mentioned on product labels.

Manufacturers avoid the term "bacteria", because it sounds icky, but you can't make dairy yogurt without it.

The mix of bacteria used varies from brand to brand; the new study focuses on one called Bifidobacterium bifidum (see photo above).

(Some brands, like Chobani and Stonyfield, use the term Bifidus on their labels. Others, like Siggi's, use the term Bifidobacterium. Same thing.)

Colorectal cancer

Colorectal cancer is the second most common cause of cancer deaths in the U.S. The incidence for both men and women under 50 has been slowly but steadily rising since 2012. It's important to know as much as we can about prevention.

A 2022 meta-analysis suggested that eating more yogurt lowers the risk of colorectal cancer. The new study attempts to better understand the connection.

Specifically, the researchers explored whether the Bifidobacterium in yogurt is the key ingredient. Their hypothesis was that by improving gut health, Bifidobacterium in yogurt lowers the incidence of tumors containing Bifidobacterium.

I realize the logic of that last sentence may seem a bit convoluted.

Here's how I would reconstruct it: Bifidobacterium is present in everyone's intestines, regardless of whether or not they eat yogurt. Bifidobacterium can also be found in some but not all colorectal tumors. If this bacterium has anti-tumor properties, then people who take in more of it via yogurt will be less likely to develop tumors that contain the bacterium. (The researchers don't explain exactly why that would happen.)

The study

The new study, published February 12 in Gut Microbes, was led by Satoko Ugai, a postdoctoral researcher in epidemiology at Harvard and a Harvard Medical School affiliate.

Data were gathered from two large, ongoing cohort projects, the Nurses' Health Study (87,379 women) and the Health Professionals Followup Study (44,677 men).

This is not the most diverse sample, but it's a good one in the sense that we'd expect health professionals to be relatively accurate at reporting their own health-related behaviors.

Beginning in 1980 (NHS) or 1986 (HPFS), participants completed lifestyle questionnaires once every four years or so. Colorectal cancer cases were identified every two years, and the medical records and tissue samples of patients were examined.

For yogurt intake, Ugai and colleagues divided participants into four groups:

• < 1 serving per month

• 1–3 servings per month

• 1 serving per week

• ≥ 2 servings per week

Main finding

The main finding was a lower incidence of Bifidobacterium-positive colon cancers among those who ate the most yogurt.

Bifidobacterium-positive cancers are tumors that contain Bifidobacterium. (About a third of colorectal cancers do.)

Based on a kind of statistic called a Hazard Ratio, the researchers calculated that people who ate the most yogurt (2 or more servings per week) were 20% less likely to develop Bifidobacterium-positive colon cancers than people who ate the least (less than one serving per month).

But don't head to the grocery story just yet.

A fatal flaw

This study is grounded in the premise that the more yogurt people eat, the more Bifidobacterium they ingest. The hypothesis and findings don't make sense otherwise.

Unfortunately, Ugai and colleagues had no data on the types of yogurt participants consumed. Rather, the researchers simply noted that yogurt "typically contains Bifidobacterium."

Guess what? It doesn't.

Yesterday I visited the largest grocery store in my vicinity, a Safeway, and examined the product labels of every yogurt sold there. (I must've looked strange, but, you know, anything for science.)

My goal was to figure out which brands contain Bifidobacterium (keeping in mind that some brands use the term bifidus).

Here's my "methodology":

––I excluded products for kids (e.g., Go-gurt) as well as non-dairy yogurts.

––I treated products with different brand names manufactured by the same company as separate products (e.g., Oikos is produced by Dannon.)

––The labels for two brands – Lucerne and Yoplait – only note that their products contain "live and active yogurt cultures", so I obtained more specific information from manufacturer and researcher websites.

––Some brands use a different mix of bacteria for different products. If even one product contained Bifidobacterium, I treated that as a "yes" for the brand.

Here's my data:

As you can see, only five of the 12 brands of yogurt sold at my Safeway contain Bifidobacterium.

Is my grocery store quirky? Does the manager have a secret bias against this microorganism?

I doubt it. Of the five most popular brands of yogurt, only one (Chobani) contains Bifidobacterium.

The researchers' statement that yogurt "typically" contains Bifidobacterium therefore seems wrong. Typically speaking, it does not.

Expert guidance

The bacteria typically found in yogurt, according to AI Overview, are Lactobacillus Bulgaricus and Streptococcus Thermophilus. (This tracks with what I observed at Safeway and on product websites.)

Since AI is known to hallucinate, I also consulted with two human experts.

First, I emailed Satoko Ugai, the lead author of the study, who was cordial but seemingly unaware of the implications of my question:

"As you mentioned, not all yogurt contains Bifidobacterium. Yogurt may contain other strains, such as Lactobacillus species. Those strains may also have anti-tumor effects."

That "may" is not very persuasive. Suppose you develop a new brand of iodized salt, give it to people to correct an iodine deficiency, find that it works, but discover later that some of your salt wasn't actually iodized. You could say that it "may" have somehow created iodine in the body anyway, but that sounds awfully implausible and biased.

Likewise, if your study is premised on the notion that Bifidobacterium in yogurt prevents Bifidobacterium-positive tumors, but not other types, then you discover that most yogurt doesn't contain that bacterium, it's not very persuasive to say, well, I guess other kinds of yogurt have the same bacterium-specific effect.

I also contacted Seockmo Ku, an expert at Texas A & M who has published several papers on Bifidobacterium. His response:

"As you correctly pointed out, not all yogurts contain Bifidobacterium...However, many U.S. food manufacturers (e.g., Chobani, Activia, Stonyfield) have recently been incorporating Bifidobacterium into their yogurt production."

Even if the use of Bifidobacterium is a growing trend, you can see from my "data" that it's not widely used just yet. Meawnwhile, the data for Ugai and colleagues' study extends all the way back to the 1980s.

Doomed from the outset

None of the 12 yogurt brands I found at Safeway rely on just one bacterial culture. The five that use Bifidobacterium combine it with other cultures. Chobani and Stonyfield, for instance, each include five others.

In short, the study never had a chance of revealing something about Bifidobacterium. Even if dairy yogurt "typically" contains it, other cultures are always present too.

So, what does the study tell us?

Alternative interpretations

Here are two alternative interpretations of the findings:

1. The results are a statistical fluke.

This isn't quite as implausible as it sounds. Although the sample consisted of 132,056 people, only 3,079 colorectal cancer cases were recorded. For statistical reasons, it's easier to find significant effects when the number of cases is relatively tiny.

2. Healthier lifestyles lower the risk of one type of colorectal cancer. Yogurt may or may not contribute.

This interpretation is consistent with the researchers' own data as well as that of other studies.

Ugai and colleagues found that participants who ate the most yogurt were also more physically active, less likely to smoke or consume processed meat, and otherwise engaged in healthier lifestyles.

Although the researchers adjusted for these differences statistically, not every health-related variable was – or could have been – measured.

In short, folks who ate the most yogurt may have had lower rates of one type of cancer owing to healthier living. Eating yogurt may have had little or no benefit.

The term "omitted variable bias" is sometimes used for the concern I've raised here. Perhaps some health variable that wasn't measured, like quality of sleep, influenced cancer rates. (It's not clear though why only Bifidobacterium-positive colon cancers would be affected.)

What should we eat?

The second interpretation is the more positive one, but it doesn't offer very specific guidance.

Rather it tells us that healthier lifestyles lower the risk of one type of colorectal cancer. In the context of this study, that means exercising regularly, avoiding tobacco, eating a balanced diet, and not overindulging in alcohol, processed meats, or red meats.

And yes, the data hints that we should eat yogurt too. But we have no way of knowing whether yogurt is essential to the mix.

I asked Dr. Ugai whether she would encourage people to switch to brands of yogurt that contain Bifidobacterium. Her response was pretty unequivocal:

"No. As I said, other strains than Bifidobacterium may have anti-tumor effects. Also, this study and other studies suggest that yogurt itself is beneficial for health, regardless of the Bifidobacterium status. More importantly, any sugar is bad for your health. I would recommend plain yogurt (not containing any sugar)."

I don't agree with every point she makes here.

The evidence that yogurt cultures other than Bifidobacterium have anti-tumor effects is far from clear. The 2022 meta-analysis that motivated the new study relies on some consistently awful research.

I do think a lot of data supports the notion that yogurt is healthful, even if Dr. Ugai's own study doesn't show it. But her claim that any sugar (meaning refined sugar, I presume) is bad for your health seems overstated and inconsistent with most dietary guidelines, which only specify daily limits.

So, what should we do? Eat yogurt?

Sure! It's a great source of protein, vitamins, and minerals. Low-fat and nonfat options are available if you're concerned about saturated fat. Low- and sugar-free options are available too if you're concerned about added sugars.

At the same time, all nutrients in yogurt are available from other sources too. You shouldn't feel that a healthy diet depends on it.

As for the bacterial cultures in yogurt, they promote gut health, but you can find them in a variety of fermented foods as well. Depending on the bacterium, your options include, among other things, pickles, sauerkraut, kefir, kombucha, kimchi, miso, and certain cheeses (cottage cheese, swiss, cheddar, provolone, etc).

Final thought

The new study has been covered by a number of national and local news outlets, including NPR and Fox News. The takeaway is that yogurt can prevent colon cancer. I think the media coverage so far has been misleading.

I love yogurt – it's tasty and nutritious – but even the most optimistic take on the data fails to show that it prevents cancer. Rather, a healthy lifestyle and regular screenings are the best approaches to prevention. Yogurt has a place in a healthy diet, but it's not an absolute necessity.

Thanks for reading!