I’m occasionally stricken by a wave of crippling, all-consuming terror. Sometimes it’s because I can’t find my wallet. Sometimes it’s because I hear the unmistakable sound of Smitty throwing up on my bed. Sometimes it’s because I take a few wrong turns on Youtube and accidentally learn what Piccinini animal-human hybrids are (what is seen cannot be unseen). But these days, it’s usually because I’ve looked at the calendar and realized that—along with being 25 and really old now—I haven’t posted anything on this blog in almost four months.
As most of you probably know, I’ve been chugging away on an upcoming book called “Death By Food Pyramid,” which is the main reason Raw Food SOS has been hosting more tumbleweeds than blog entries lately. Thanks to finding some unexpected political shenanigans to investigate (which I’m really excited to tell you guys about), the release date for “Death By Food Pyramid” is now September 2013. More details to come.
Earlier this month, I recorded an interview that touches upon the USDA’s seamy, pyramid-shaped underbelly (mostly in the second half):
I’ll be writing more about the book soon (and resuming my previous rapid-blogging schedule of six posts a year instead of four), but in the meantime, here’s a new installment of Bad Science Du Jour!
Are low-carb diets killing Sweden?
Last week, I wrote a guest post for Mark’s Daily Apple on a study that seemed to link Sweden’s “Low Carb High Fat” (LCHF) movement with an increased risk of heart disease. As often happens when media outlets are vying for readership, the headlines traded accuracy for alarmism—and it turned out the study had more to do with cheese-puffs and Justin Bieber than heart disease and carb restriction. You can read my breakdown of it here:
But it seems the Peer-Reviewing Powers That Be are smiling upon us this June, because Tuesday marked the arrival of yet another study on low-carbing Swedes—this one with a bit more meat on its bones. The BMJ paper (readable here until a paywall slurps it up) implied a grim future for carb-cutting ladyfolk. As Huffington Post UK reported (and you know they’re legit),
Women on Atkins-style diets are putting themselves at risk of heart disease and strokes, experts have warned. Those who regularly eat a low-carbohydrate, high-protein diet are at greater risk of cardiovascular disease than those who do not participate in such diets.
What’s the story, Wishbone? Do you think it’s worth a look?
This is a great study to tweeze apart for a few reasons, hence why it lured me out of blogging silence—but first, an important disclaimer:
I don’t think low-carb diets are universally flawless, especially if food quality isn’t emphasized—and if a long-term study emerges that shows legitimate concerns, I’ll be the first to tell you about it (or maybe the 10th, given my blogging speed). I deeply respect the success many folks have on low-carb regimens, and I have no doubt that this way of eating is often a lifesaver—sometimes quite literally—for folks facing obesity or certain health conditions, or who found themselves trapped in a low-fat-diet-hunger death spiral. At the same time, I think it’s clear by now that some people do well on the carbier end of the spectrum (myself included) and may run into problems on very-low-carb menus. And that’s okay! We’re all special snowflakes.
What I’m saying here is that I like to blog in defense of low carb not because I think it’s the holy grail of macronutrient ratios or the only healthy way to eat, but because it’s the bullied Poindexter of the diet world: no matter what its studies reveal, the media can’t help but punch it in the face, steal its lunch money, and drench it in buckets of pig blood à la Carrie. Few other diets receive such nasty—and generally unfounded—backlash. If we’re gonna embrace this whole “science” thing, we need to fight hard for objectivity wherever it may be lacking. En garde, y’all!
That said, I’ll spoil this study’s ending right off the bat and say that 1) it was observational and 2) it didn’t actually follow low-carbers. Yes, double whammy of ughness. We could technically stop right there and turn our thoughts to more important matters, like why traumatizing things exist on the internet (Piccinini hybrids, I’m looking at you), but I think that’d be a missed opportunity. This is actually a valuable study to analyze because it uses a “diet score” design that’s quite popular—and becoming ever more so—with observational health research. (I blogged about one such study in 2010, but there exists a wealth of other diet-score studies floating in the research sea, woefully undissected.)
Thusly, understanding this week’s study will help us understand the future ones—inevitably incubating in some researcher’s mind as we speak—that will one day hit the news with the same design, the same conclusions, and the same slew of limitations.
(Note before we kick things off: the following who-knows-how-many paragraphs are excruciatingly exhaustive, probably more than they need to be. My goal is to explain the details of this study in a way that makes this dissection relevant for future studies of a similar design. If you just want a summary of why the study is bad instead of my long-winded breakdown (no offense taken!), check out the study’s comment section on the BMJ website, where readers have pointed out many of the paper’s flaws in fewer words than I’ll be using.)
The design… of doom
In the early ’90s, about 50,000 Swedish females—all between the ages of 30 and 49—filled out a diet-and-lifestyle survey as part of the Women’s Lifestyle and Health Cohort. The women self-reported (by memory) everything they’d eaten during the past six months, answered some lifestyle questions, mailed their survey back to the project leaders, and then went on their merry way. This was the only time women reported their food intake during the entire 16-year study.
After kicking some ladies off the data-island because their questionnaires were incomplete (or because they entered the study already diagnosed with cardiovascular disease), our researchers were left with 44,000 participants. They translated each woman’s food-frequency responses into daily intakes for protein, carbohydrate, fat, and total calories, and then arranged that new macronutrient data into deciles—10 equal groups divided at the 10th percentile of intake, 20th percentile, 30th percentile, 40th percentile, and so forth. In the case of protein, for example, the first decile would include data from women with the very lowest intakes of protein, while the 10th decile would include data from women with the very highest intakes of protein. (Another way to look at it, since this system is based on tens: if a woman falls in the first decile for protein, nine deciles’ worth of women—or 90% of the study population—are eating more protein than her. If a woman falls in the 6th decile for protein, only four deciles—or 40% of the population—are eating more protein than her.)
Lather, rinse, and repeat this decile system for carbohydrate intake, fat intake, and total calorie intake.
Isn’t math grand?
But we’re just getting started. Here’s where the “diet score” stuff enters the scene.
Diet patterns: the easiest way for a scientist to score
In order to extort information about low-carb diets from a non-low-carbing population (an ambitious feat, to say the least), the researchers created a “low carbohydrate, high protein” diet score, and assigned points to each woman based on which carbohydrate and protein deciles she landed in. For protein, women would get one point if they were in the first decile; two points if they were in the second decile; three points if they were in the third decile—all the way up to 10 points for falling in the 10th decile. The point-assigning system was inverse for carbohydrate: women would get 10 points if they were in the first decile, nine points in the second decile, eight points in the third decile—on and on—and one point if they were in the 10th decile.
Once those calculations were made, the researchers tallied up each woman’s protein points and carbohydrate points for her combined “low carbohydrate, high protein” score, which could range from 2 to 20.
The purpose of that score was to gauge how closely each woman’s diet resembled the low-carb, high-protein eating pattern the researchers wanted to study. Women with the lowest scores were the furthest away from that pattern; women with the highest scores were, rightly or wrongly, judged to be Atkins fangirls. Some score examples could go like this:
Woman eating lots of fruits, vegetables, and grains, but very little animal food (i.e. Campbell’s future BFF): 1st decile for protein, 10th decile for carbohydrate = 1 + 1 = score of 2
Woman eating lots of lean meats, skim milk, very little fat, and moderate grains and potatoes: 5th decile for protein, 4th decile for carbohydrate = 5 + 7 = score of 12
Woman eating a very low-carb, low-protein diet with abundant fats, organ meats, butter, and cream: 1st decile for carbohydrate, 2nd decile for protein = 10 + 2 = score of 12
Woman eating a very low-carb, high-protein diet with lean meats, egg whites, fish, and protein powder: 1st decile for carbohydrate, 10th decile for protein = 10 + 10 = score of 20
(If you spot the problem here, fear not—we’ll come back to this in a moment!)
Armed with their meticulously-calculated diet scores, the researchers embarked on their next mission: to see which women received a “cardiovascular disease” diagnosis between 1991 and the end of 2007, and—the study’s crux—how their diet in ’91 influenced their health over the next 16 years.
At last, the moment of truth.
After running a battery of sophisticated analyses—looking at cardiovascular disease both collectively and in subcategories, and adjusting for common confounders like smoking, body mass index, education, and exercise—the researchers found the same thing most studies of this kind uncover: that women with the highest “low carb, high protein” scores had a higher incidence of cardiovascular disease than women with the lowest scores.
Straight from the paper (emphasis mine):
We found that women had a statistically significant 5% increase in the incidence of cardiovascular events per 2 unit increase in the 20 unit low carbohydrate-high protein score. In practical terms, and taking into account the rough correspondence in the ranking of energy adjusted and crude tenths of intake, a 20 g decrease in daily carbohydrate intake and a 5 g increase in daily protein intake would correspond to a 5% increase in the overall risk of cardiovascular disease.
Although a 5% risk increase might not sound like much at first, keep in mind that it applies to each 20-gram decrease in carbohydrate intake—supposedly adding up to a much bigger risk for true low-carb adherents. No wonder the media gobbled this one up!
Alas, this study is the last place to look for anything informative about low-carbohydrate diets—or really any diets, for that matter. Although the researchers did a commendable job analyzing the data they had, the rest of this study is a hot mess of defects. You’ve probably spotted some of them already, but a few important problems are tucked out of immediate sight. And we’re about to blow them all open.
Let’s start at the very beginning: the Swedish diet survey from the early ’90s.
Food frequency questionnaires… of doom
If you haven’t read my Will Eating Red Meat Kill You? guest post on Mark’s Daily Apple from a few months ago (which I can’t hold against you, since I never posted the link!), I recommend hopping over there right now to skim the section on food frequency questionnaires. (If two long blog posts are too much Denise for one day, check out this classic piece by Chris Masterjohn on the unreliability of self-reported data instead.)
The bottom line—discussed more thoroughly in those aforementioned links—is that bad things happen when you ask people to report what they eat. That badness only amplifies when they’re reporting from memory, and becomes incrementally worse the further back in time you want them to recall. Most folks just aren’t that aware of what goes in their mouths, especially when those things went in their mouths six months ago. As a result, nearly everyone underestimates what they really consume, and some foods—usually the ones we think are bad for us—tend to be a major roulette-wheel spin in terms of accurate reporting. Health-conscious folk may be particularly likely to underreport their intake of “bad” foods out of sheer guilt (who wants those Saturday brownie binges emblazoned forever on paper?). And since not everyone misreports their food intake in a consistent way, no amount of statistical wand-waving can really make up for the inaccuracies in this type of data.
And if that hasn’t convinced you of food questionnaires’ inherent dodginess, maybe this will. Here’s the actual survey those 44,000 women of this study had to fill out in the early ’90s, when they first enrolled in the Women’s Lifestyle and Health Cohort. The diet questions start on page 13.
Here’s a prime example of the torture contained within:
And that’s only the first page. The full table includes 49 more items, just in case the ones above were too easy. I have to wonder how many people saw this questionnaire and were like,
And then filled it out like,
My guess is more than a few.
Indeed, the data from this survey seems pretty sketchy—both in terms of accuracy and in terms of being anything even remotely usable for “low carb” research. Even the validation study for the questionnaire, which was conducted to test how well the self-reported data matched direct diet records (where participants actually weighed and recorded their food each day), notes that refined grains, sweets, cooked and fried potatoes, vegetables, processed meat, poultry, fruit, and whole grains were over- or under-reported by at least 20%!
To illustrate the problem further, this table summarizes the diet data for the women in our current study:
Check out that calorie range! Based on the researchers’ translations of the survey data into dietary intakes, the women were eating 4241 kJ (only 1013 calories!) in the 10th percentile and 9053 kJ (2162 calories) in the 90th percentile. Think about that 1000-calories-a-day figure. Were thousands of ladies inexplicably eating starvation-level diets for months on end? Or were they, perhaps, just doing an unsurprisingly poor job of remembering how many broccoli florets, almonds, and tablespoons of salad dressing they’d eaten in the past year?
I’ll let you be the judge.
Equally damning is the data for macronutrients. The cutoff for the first decile of carbohydrate intake (the column marked “10th centile” in this table) is 123.7 grams a day—moderate carb at best, and far higher than most legitimately low-carb eating plans allow. Yet this, ladies and gentlemen, is the decile representing the lowest of the low carbers in our study. Women eating up to 123.7 g of carbohydrate per day received the full 10 “low carb” points when the researchers tallied up their scores. And since many women whose carbohydrate intake landed in the first decile are probably the same ones allegedly eating only 1000 calories a day, their true carbohydrate intake is likely much higher, making this already-not-a-low-carb-study even more so.
Macronutrient warz… of doom
By now, you’ve probably noticed one of this study’s many quirks: the researchers assume that carbohydrates and protein are hinged together on some sort of macronutrient seesaw—where reducing carb intake automatically means increasing protein intake, and vice versa. Although some low-carb plans do emphasize protein, it’s far more common for fat to be the nutrient that fills the calorie void when carbs go down, with protein changing to a much lesser degree. That’s because the vast majority of foods are either carbohydrate dominant (like grains, fruits, and starchy vegetables) or packed with plenty of fat (like meat, non-adulterated dairy products, and nuts), with relatively few whole foods—mostly shellfish and some non-fatty fish—being straight-up protein bombs. Unless you go wild with egg whites and protein powder and don’t mind meals that taste like nothing, it’s pretty hard to raise your protein intake particularly high, at least relative to fat and carbohydrate.
So why did the researchers fixate on a low-carb, high-protein diet score instead of a low-carb, high-fat one? Your guess is as good as mine. Perplexingly, they write in their paper that “The most popular among [weight-loss] diets emphasise reduction of carbohydrate intake, thereby encouraging high protein intake, as high fat diets are generally avoided in most Western societies.” Seems fair enough. But the paper they cite for that statement starts off by describing low carb as unabashedly fat-full:
The diet claims to be effective at producing weight loss despite ad-libitum consumption of fatty meat, butter, and other high-fat dairy products, restricting only the intake of carbohydrates to under 30 g a day.
It’s a head scratcher, that one.
As it stands, in this study, a true low-carbing woman with a moderate protein intake would only get a midrange score—and could easily end up with the same number of points as someone eating lots of carbs but also lots of protein. The bizarre “low carb, high protein” scoring system ultimately does a poor job of identifying any true low carbers in the cohort.
And best of all: towards the end of their paper, the authors eventually acknowledge that the “low carb, high fat” diet pattern in this study was a far cry from what actual low-carb programs promote (emphasis mine… as always):
Among the women studied, carbohydrate intake at the low extreme of the distribution was higher and protein intake at the high extreme of the distribution was lower than the respective intakes prescribed by many weight control diets. However, the underlying trend between low carbohydrate-high protein score and incidence of cardiovascular disease was essentially monotonic, indicating that our findings are applicable across the spectrum of carbohydrate and protein intakes of the participating women.
Translation: “We realize our low-carb study wasn’t actually a low-carb study, but let’s not let that prevent us from using it as evidence that eating low carb is bad.”
What about fat? Could that be the real reason the lowest-carb women got more heart disease?
Now here’s where things get interesting.
In the paper’s Results section, we see a brief (but important) note about fat: “The additive low carbohydrate-high protein score was … positively correlated with lipid intake (Spearman r=0.28 for saturated lipids; Spearman r=0.17 for unsaturated lipids)”—which means that as women progressed upwards on the low-carb, high-protein scale, their intake of fat was generally going up too. No brainer, considering high-protein foods often have plenty of fat.
But the researchers were apparently so committed to making this study only about low-carb, high-protein diets that they did something quite surprising: they designed their statistical models to adjust for both saturated and unsaturated fat intake (in the same way they adjusted for smoking, BMI, and so forth), to completely remove the influence of all things lipid! In other words, the results of this study can’t arguably be linked to any form of dietary fat, because the researchers statistically removed its effects. Whatever was driving the women’s cardiovascular disease risk was decidedly non-greasy.
Presumably, the researchers figured saturated fat would contribute to cardiovascular disease and thus treated it as a confounder. But since saturated fat is—at least by mainstream wisdom—deemed a major reason for the unhealthfulness of low-carb diets, it seems odd that the researchers totally removed it from the equation. What’s left to take the blame for the increased cardiovascular disease? The lowness of the carbohydrates? The highness of the protein?
The researchers offer a pretty standard explanation:
With respect to the biomedical plausibility of our findings, vegetables, fruits, cereals, and legumes, which have been found in several studies to be core components of healthy dietary patterns, are important sources of carbohydrates, so that reduced intake of these food groups is likely to have adverse effects on cardiovascular health.
(I probably don’t need to point out the obvious—that many low carbers end up eating more vegetables than their non-dieting peers, and that no heart-healthy nutrient exists in grains that can’t be obtained elsewhere—but given that the lowest low carbers in this study were still eating up to 123 grams of carbohydrate a day (and likely much more due to underreporting), I doubt they were suffering en masse from a grain-and-legume deficiency.)
Incidence rate ratio… of non-doom
Now that we’ve determined that this study 1) draws on terribly unreliable self-reported data, 2) doesn’t examine diets in a genuine low-carb threshold, and 3) has nothing to do with saturated fat, let’s take a look at just how risky the “low carb, high protein” pattern really was for cardiovascular health.
The researchers—and the media outlets that reported on this study—describe the results as showing that for every 20-gram decrease of carbohydrate intake (and every 5-gram increase in protein), the women’s cardiovascular disease risk rose by 5%. Less publicized was that when the researchers used women with a “low carbohydrate, high protein” score of six or less as their baseline, they found that women with scores above 16 had an incidence rate ratio of 1.60—or a 60% increase in risk compared to their carb-loving, protein-shunning counterparts.
But numbers are a tricky, sneaky thing when they’re relative. Here’s a table showing the actual quantity of cardiovascular disease diagnoses for various score categories (the “incidence rate” is per 10,000 woman years):
Let’s focus on that All cardiovascular disease category. The researchers note that “potential confounding influences have not been accounted for” in this table, so the incidence numbers—especially in the highest- and lowest-point categories—are likely reflecting influences other than diet. But we’ll let that slide for now.
From this table, we can see that for women with a score of six or less, 2.2%—or 22 out of every 1000—were diagnosed with cardiovascular disease during the course of the study (188 incident cases divided by 8343 women). For women with a score of 16 or higher, 3.6%—or 36 out of every 1000—were diagnosed with cardiovascular disease (302 incident cases divided by 8395 women). That’s an absolute increase of only 1.4%, even though it represents a scarier-sounding relative increase of about 60%. And considering the 1.4% is a confounder-riddled value likely to over-represent the actual difference, that number becomes even less impressive.
But that’s not all. Let’s look again at the incidence rate ratios the researchers (and beloved media) gave us earlier: 1.05 for every two-point increase in the “low carb, high protein” score (or for every 20-gram drop in carbohydrate intake/5-gram rise in protein intake). Is that really a noteworthy ratio?
Lo and behold, the researchers help us answer that very question. In a table too unwieldy to capture in a screenshot, they list the incidence rate ratios (age-adjusted) for all cardiovascular disease diagnoses, based on a variety of characteristics unrelated to food:
- Being a current smoker (versus no smoking history): 2.78 — a 178% risk increase
- Having a body mass index in the “obese” range (versus in the normal range): 2.48 — a 148% risk increase
- Having a body mass index in the “underweight” range (versus the normal range): 2.17 — a 117% risk increase
- Having a hypertension diagnosis (versus no hypertension): 2.50 — a 150% risk increase
- Having the highest self-rated level of physical activity (versus the lowest): 0.37 — a 63% risk decrease
- Having over 13 years of education (versus 10 or less): 0.45 — a 55% risk decrease
- Being over 170 cm in height (67 inches, or 5’7″) (versus being 160 cm/63 inches or under): 0.77 — a 23% risk decrease
Compared with things like going to school for a few extra years and not being vertically challenged, the risk associated with the “low carb, high protein” diet score is minuscule!
But frankly, it hardly matters anyway. All these numbers are based on a single survey from 1991 – ’92, with absolutely no follow-up questionnaires to see if the women’s diets or lifestyles changed afterwards. This whole study rests on the assumption that 44,000 women not only accurately reported their yearly food intake in the early ’90s, but that they kept eating, exercising, and weighing the same for the next 16 years. (Or conversely, that whatever they ate in 1991 singlehandedly determined the fate of their health for the rest of their lives.)
To sum it all up…
- This study is observational—meaning it tells us nothing about cause and effect.
- This study relies on one food frequency questionnaire that the women filled out upon enrollment in the Women’s Lifestyle and Health Cohort. We have no idea what they were eating during the next 16 years.
- Women reported their food intake during the early ’90s, when low-fat diets were the gold standard for healthy eating. The most health-conscious women were likely to be eating the most carbohydrate at that time.
- Food frequency questionnaires are all sorts of terrible and don’t let anyone ever tell you otherwise.
- The study’s “diet score” design assumes that all low-carbohydrate diets are necessarily high in protein—resulting in a point-assigning system that doesn’t always identify true low carbers.
- The women with the lowest carbohydrate intake in this study were still eating up to 123 grams of it a day (an amount that’s probably wildly underestimated, to boot)—meaning they weren’t actually eating a low-carbohydrate diet.
- The true increase in cardiovascular disease risk from scoring high on the “low carb, high protein” scale is pretty minor—and pales in comparison to other lifestyle variables.
If your eyes are glazing over and all you care about is this blog entry being over, just remember this one thing:
Good math can’t make up for bad data.
This applies to every single study in the whole world that relies on food frequency questionnaires. These studies are not your friends. It doesn’t matter whether they yield results that make you feel better about your chosen lifestyle or worse for it—they’re all built on shaky ground, and not even the finest statisticians can change that fact. In order to have reliable results, a study needs reliable data. And for the most part, food-questionnaire-dependent studies are not the place to find it.
Likewise, any type of “diet score” study using data from the general population will suffer from one fatal flaw: these studies try to squeeze information about specific, clearly defined diets from people who—by and large—don’t actually eat those diets. It’s sort of the nutritional equivalent of flying to Paraguay to study Greek architecture. The reason this type of study design is popular isn’t because it has many redeeming qualities, but because it’s a heck of a lot easier (and cheaper!) to fiddle around with pre-existing data from large cohorts than to run controlled experiments where you make people do what you want them to do.
So basically, this study tells us nothing about nothing. Most other studies like it also tell us nothing about nothing. What does that leave us with?
Something like this, I’d say:
Although observational studies are a bucket of fun to peel apart, they’re also the sitting ducks of the research world—easy targets, limited by design, and utterly feeble behind their smokescreens of media hype. There’s only so much you can do with ’em before the study-dissection process becomes redundant and you start feeling bad about attacking something too incapacitated to fight back. When I’m back here blogging more regularly, I’ll be covering a great many topics other than observational studies—I promise!
…What, you thought this was over? I’m not letting you go that easy!
What follows is an assortment of things that don’t really deserve their own blog posts but that I want to mention somewhere, so I’m just going to tack them awkwardly onto the end of this entry and hope you guys don’t notice how weird it is.
1. For anyone who hasn’t seen it yet, I added a “For Vegans” page at the top navigation bar with some health tips for committed vegans. Link here:
2. A few months ago, I wrote a summary of the major flaws in “The China Study”—mostly based on my Wise Traditions presentation from last year—that you can read here:
3. For organizational purposes, here are my two most recent guest posts at Mark’s Daily Apple that I forgot to link to from this blog:
4. Are raw vegans immortal? Once in a while, I like to peruse the search-terms page on WordPress to see how wayfaring Googlers find my blog. After the standard pileup of “China Study” and “Forks Over Knives” phrases and various creative spellings of my name, things start getting weird. This came up not too long ago:
And then there’s tons of this:
Although I can’t honestly say I know why someone “loves popcorn craves” (or what that even means), I can say that vegans do get cancer, and raw vegans definitely die. Sometimes from cancer, no less. I think the general vegan community is pretty realistic about the fact that becoming a vegan—even the junk-food-shunning, non-smoking, non-drinking, non-drugging, non-rabid-honeybadger-owning variety—isn’t the same as putting on an invincibility cloak and forever evading disease, but raw vegans often have more confidence in their eternal immunity, which I find concerning. Case in point: the responses to this post from 30 Bananas a Day.
At any rate, the can-vegans-get-cancer question seems to be the most common of the bunch, and I’m not surprised. I’ll wager that “The China Study,” especially Campbell’s research with casein, has left some folks with the impression that—even if you’re exposed to a nasty carcinogen like aflatoxin—cancer can only happen if you’re eating enough animal protein to feed it. Raw vegans often feel they’ve got added protection because they’re eating nothing but quintessential health foods.
At the risk of being kind of morbid, I want to put up a list of some really-long-term raw vegans (and a few non-raw ones) who got diseases they weren’t supposed to get. I don’t think this is common knowledge—I’ve only gathered this list piecemeal from hanging out in the vegan gossip stream for much of my past—so here it is for the rest of the world to see.
- Bif Naked—award-winning musician and actress—was a raw vegan for 10 years before getting diagnosed with breast cancer.
- Ross Horne, famous Natural Hygienist and author of a book called “Cancer-Proof Your Body,” died of prostate cancer in 2004—after 22 years of eating a high-fruit, raw vegan diet in the tropics. (Although the previous link says he died “well into his 80s,” his actual age at death was plain ol’ 80.)
- Harvey Milstein—leader of the raw vegan “natural hygiene” movement in Houston, Texas—died of colon cancer in his 60s after 23 years as a fruit-based raw vegan.
- Vihara Youkta, dancer and wife of famous raw foodist Viktoras Kulvinskas, died of bowel cancer after being raw vegan for 30 years.
- More recently, Robin Gibb of Bee Gees fame died of colon cancer at the age of 62, after being a drug-and-alcohol-abstaining vegan for decades (cooked, not raw).
- T.C. Fry—fruitarian and early raw vegan pioneer—died of a coronary embolism when he was about 70, and his autopsy revealed additional blood clots in his legs. (Joel Fuhrman notes that Fry’s B-12 levels were extremely low before his death, so he probably had high homocysteine, which makes blood more prone to clotting.)
- And on the cooked side of things, we’ve got H. Jay Dinshah—the founder and former president of the American Vegan Society—who died at age 66 of a heart attack. He was a vegetarian from birth and vegan for his final 43 years of life. (Here’s an excellent write-up about the implications of his death by Michael Klaper.)
There are definitely some long-lived cooked vegans (like Donald Watson, who invented the word “vegan,” remained one for 60 years, and died peacefully at the age of 95). But so far, raw veganism has had a less-than-impressive track record for longevity. The raw foodists who’ve made it into old age in good health typically included some animal products. Fifty-year raw foodist Norman Walker—who lived to be 99 (despite rumors of a supercentenarian status)—used goat’s milk, raw cream, and cottage cheese. Bernando LaPallo, a still-living 109-year-old, mostly-raw foodist from Arizona, eats fish along with his fresh fruits and veggies.
Hopefully people who google “can vegans get cancer?” will now find this page.
Now I’ll let you guys go. Here’s a picture of some baby hedgehogs so we end on a happier note.