Get cancer[1] and you’ll be inundated with advice about food, most wrong and much contradictory: avoid sugar; processed food kills; meat promotes cancer; ketogenic diets are incompatible with cancer; milk is dangerous; milk is healing; cancer is impeded by vegetables; honey is good; tea is safe; coffee is dangerous; tumors like Adderall. I think people latch onto diet-based advice, like they do prayer, as a mechanism of control, even if the mechanism is faulty, in a situation where any control is highly medical and scientific and thus beyond the typical person’s abilities. Today, the most common form of nutrition advice is to avoid sugar, even though no evidence suggests that a low-sugar diet will eliminate or reduce cancer (“cutting out all sugars doesn’t actually fight existing tumors”)[2]. Sure, it’d be nice if one could follow a particular diet advice in order to eliminate tumors. “It’d be nice” is not the same as “it is true.”

There’s a human tendency to crave control even if craving control isn’t well correlated with true control over a chaotic, uncertain world. One sees this, for example, in people who erroneously think driving is safer than flying—a driver has some control, but, on a per-mile basis, flying is way safer. We imagine that that we’ll evade the drunk driver, the woman distracted by her smartphone, the dude yelling at his kids. In reality, we don’t control the cars and distracted drivers around us. By contrast, no amateurs fly large planes, the pilots are sober experts, and the FAA extensively digs into any crash to figure out how to prevent the same thing from happening again.

We collectively (and bizarrely, in my view) accept 40,000+ car fatalities every year in the U.S. alone. To me this is insane, but I’m the weirdo in that most people don’t think statistically and accept the fact that people they know and sometimes will be seriously hurt or killed in car crashes. In Sweden, however, there’s an effort to understand the factors underlying serious car crashes, and, because of that effort, “Today, Sweden has some of the lowest rates of road traffic fatalities in the world.” In the U.S. and much of the world, we tend to blame individual drivers, instead of systems; systems, however, can (and should) be improved. In Sweden, “officials were no longer allowed to design roads for idealised drivers who never became distracted or exceeded the speed limit. They had to make roads for real people who made mistakes.” Drivers still drive, but the roads are built to limit the ability of people to kill and maim one another. Real control happens at the level of the system.

I’m not saying striving for control is bad; given the frantic, relentless efforts Bess and I have put into keeping me alive, it’d be peculiar and hypocritical if I did. Striving for greater control is good, but grasping at illusions is not, particularly if those illusions are deleterious to the desired outcome, like, for me, staying alive via clinical trials. Individuals influence their cancer risks; not smoking, or not smoking much, is an obvious way. And it is true that high sugar intake is a risk factor for developing cancer. Once someone has cancer, though, eating sugar (or not), or eating meat (or not) isn’t going to affect the cancer’s course. What will is the usual: surgery, chemotherapy, radiation therapy, oncology treatments (like the bispecific antibody I got, or the antibody drug conjugate (ADC) I’m on now). I’ve been on a low-sugar diet for more than a decade, and that didn’t stop a squamous cell carcinoma from growing in my tongue. There’s a correlation between a low-sugar diet and avoiding cancer, but it’s far from r = 1.0.

I wrote previously:

I’m reading a biography of Richard Feynman, whose first wife died of tuberculosis (TB) in part because the scientific / medical establishment wasn’t able to get its act together regarding antibiotics: “the first clinical trial of streptomycin” began with only two patients in the fall of 1944, despite TB being a death sentence. And it wasn’t until August 1945 “that the Mayo trial had expanded to as many as thirty patients.” Finally, “In 1947 streptomycin was released to the public”—two years too late for Arline Feynman.

James Gleick, the author of the Feynman bio, quotes Feynman writing to Arline as she’s dying: “Drink some milk!” But Arline didn’t need milk; she needed streptomycin. Today, telling someone with a bacterial infection to “drink some milk” or “avoid sugar” or “try the keto diet” would be widely, justifiably laughed at.[3] “You have MRSA—drink some milk, 50ml of honey, and eat some wild boar.” Yet people commonly do this with cancer, maybe because it’s too hard for some people to say: “This sucks, and I hope expert treatment succeeds.” We live in an age in which skepticism towards experts is sometimes justified—if I’d listened to Dr. Savvides, who was once assigned to be my medical oncologist at Mayo Phoenix, I’d have done some palliative chemotherapy and rapidly died—but is often not. The trick is distinguishing between when evidence creates reasonable doubt regarding expert judgment, versus the more frequent occasions when  expert judgment is in fact correct.[4] The people who come to adopt skepticism and contrarianism as an identity often fail to distinguish the two situations.   

I wonder if there’s an element of blame, or the desire to find blame, underlying these dynamics. I’ve been quizzed by friends and strangers about all kinds of personal choices, including diet, but also exposure to radiation, or “toxins” (never defined), or smoking, or cell phones—as if by pinpointing some flaw in my own life will help the person asking protect him or herself. It can’t be random—it has to be something I’ve chosen. People did this with COVID, too, and Bess reports that, at the height of the early pandemic in NYC, when discussing the patient deaths she saw in the ER, many friends’ first question was about the habits of the deceased: were they smokers? They were smokers, right? Asthmatics? Obese? The subtext being that if the answer was “yes,” the person asking the question—who was typically a thin non-smoker without lung disease—would be exempt from negative outcomes. If Bess responded “no,” the person would frantically look for other reasons that might exclude them from being one of the future, unlucky dead. Later, as knowledge about long COVID spread, many people sought reasons they’d not be among the sufferers.

I have a similar effect on people, as if I’m a voodoo talisman of bad luck. When they find out that I have R / M HNSCC, originally of the tongue, they ask if I was a smoker, a tobacco chewer, a heavy drinker, a mime.[5] Upon telling them that I was just unlucky, I see a pall settle. They, then, could be unlucky too. But maybe not if they do that juice cleanse advertised by a thin, pretty girl on Instagram? Out comes the wallet, a case of celery and kale juice arriving like a ward against the evil eye.

It’s not just “dumb people” or “statistically illiterate people” who think diet cures might succeed. Steve Jobs’ pancreatic cancer may have been curable when it was first discovered, but instead Jobs decided to eat fruit:

Jobs initially resisted conventional medical treatment, saying, “I didn’t want my body to be opened … I didn’t want to be violated in that way.” He believed in the power of his diet and “magical thinking” to heal himself. “I think that he kind of felt that if you ignore something, if you don’t want something to exist, you can have magical thinking. And it had worked for him in the past,” Isaacson wrote.

I mean, I didn’t want my body to be violated via the first surgery that removed the initial carcinoma in my tongue, but surgery seemed better than dying and not waiting too long to get surgery would’ve help to further prevent that outcome. Jobs’s diet probably didn’t hurt him, or hurt him much if it did at all, but he needed a surgeon, not a fruitmonger:

His delay in undergoing surgery, as recommended by doctors, was a decision he later regretted. Isaacson told CBS News, “We talked about this a lot. He wanted to talk about it, how he regretted it. … I think he felt he should have been operated on sooner.”

I have made many mistakes, but not that one.[6]

It’s good for facts to replace superstition, though I don’t ever expect that to happen on a complete, widespread scale. Humans are correlation machines and many correlations we notice are spurious. True cause and effect are so often hard to distinguish. We need principles and institutions to help us be better thinkers. On our own, we’re too prone to too many biases, and we lack the intellectual tools necessary to distinguish truth from falsehood. We look back at people who thought witches caused bad weather or crop failures and pity them for their ignorance. When we really get cancer under control through mechanisms like personalized mRNA vaccines, I expect we’ll look back at the people who counseled dietary changes for cancer the same way we look at people who scapegoated witches. We’ll probably look at most contemporary medical treatments the way we look at medicine from the Dark Ages: wrongheaded, violent, and pityingly tortuous.

Hippocrates might’ve said, “Let food by thy medicine and medicine be thy food,” but he didn’t have access to chemotherapy, radiotherapy, or antibiotics.

Oh, and I’ve been drinking a lot of milk, to try to keep my weight up. It lets me eat spicier food, too.

If you’ve gotten this far, consider the Go Fund Me that’s funding ongoing care.

---

[1] I don’t recommend it, but you’ll see some interesting aspects of human social life and psychology if you do.

[2] Some studies show that high blood sugar actually makes patients respond better to chemotherapy. They’re far from the final word, however.

[3] Based on increasingly frequent internet medical advice recommending caustic essential oils to treat any manner of ills for which we have high-quality evidence-based medical treatments, recommendations like this need to be more widely laughed at as a public health effort.

[4] “Trust, but verify.” Oncologists, including ones who run clinical trials, know a lot more about oncology than I do, but Bess and I often know more about the totality of the clinical trials available for R / M HNSCC than most individual oncologists do. A surprising number haven’t heard of petosemtamab / MCLA-158, despite it having a disease-control rate around 70% (astoundingly good for R / M HNSCC) and it subsequently getting breakthrough therapy designation from the FDA. That petosemtamab isn’t already approved for R / M HNSCC is in my view criminal. The FDA is leaving people to die.

[5] They don’t actually ask about the mime thing, but you are paying attention, right?

[6] When I get dubious advice in person I tend to nod along and then change the subject; when it’s left as comments here on The Story’s Story, I delete them. I recognize that ineffective and possibly harmful people are nonetheless well-meaning, however poor their epistemology.