George Zaidan is a chemist who does not appear to have a confirmed bias for or against processed food. He views product ingredients on the basis of chemistry. He is also a communicator with an unusual writing style—not one I particularly enjoy. It might be my age. His irreverent approach disguises a philosophical message. If nothing else, the book provides a step-by-step guide to critical thinking. I confess that his analysis supports my confirmation bias. I challenge any critics of ultra-processed food to see if and where they can pick holes in his reasoning. Zaidan uses the standard Cheeto as his prime example of an ultra-processed food. As usual in my book reviews I respond to his comments in bold:
“One popular definition of ‘processed food’ is based on how complicated the food seems. This boils down to two things: how many ingredients it has and how pronounceable these ingredients are.” (p.7) Michael Pollan warns us of the dangers of unpronounceable ingredients in Food Rules. Carlos Monteiro codifies the concept in the NOVA classification of foods. Zaidan declares that food is a complex combination of chemicals. Most of these components are natural chemicals that are not required on a product label. Fruits and vegetables are much more chemically complex than his standard Cheeto. We both agree that all food is chemical, but he cautions me that nobody is listening to us when we talk like that. I may need to change my ways!
“Packaged guacamole is ‘high-pressure processed,’ meaning that we squeeze the living crap out of it, which kills microbes and also inactivates the enzyme responsible for turning the guacamole black.” (p. 72) High-pressure processing is one of the few ‘food processes’ that we can’t do at home. Processes we perform in the kitchen include baking, canning, drying, freezing, and frying. Other high-pressure products feature hummus suitable for dipping a baby carrot or Cheeto. The difference between processing food in the home and in a processing plant is usually the amount of food made and the ingredients used. The types of processes used in the home and in the plant do not differ that much. Critics of ultra-processed products claim that specialized ingredients used to make them are cheaper and less safe. Food chemists reject the premise that these ingredients are less safe.
“Most of the questions that concern our health these days have nothing to do with scurvy or pellagra and everything to do with heart disease, cancer, diabetes, Alzheimer’s, and other chronic diseases.” (p. 170) Scurvy and pellagra threatened American health during the Great Depression. Public health officials favored vitamin fortification of breads and breakfast cereals. America defeated these deficiency diseases through fortification. This success has now led some writers to conclude that if we don’t have a classical disease then we don’t have a deficiency. I contend that we need higher doses of vitamin C and niacin than needed to prevent scurvy and pellagra. Let’s eat a diet rich in vitamins and minerals without overdosing on those found in bottles on the shelf.
It is no longer acceptable to blame people for becoming obese or chronically ill. Blame has now shifted to the foods we eat. Food additives, most of which have been part of our diets for years, have now become the scapegoat. Advocates to remove additives from foods call products with these ingredients ultra-processed foods. Then they declare all ultra-processed foods unhealthy. Blaming the person or blaming the food for disordered eating misses the point. Could our diet culture, which may lead to a food obsession, be to blame for our condition? I do not believe that blaming the person or the food for the development of chronic diseases is the answer.
“Because single foods were miracle cures for horrific nutritional diseases, we were ready to accept the idea that single foods could be miracle cures for heart disease and cancer.” (p. 171) Too often in conflict and nutrition, we fight the last war instead of facing new challenges. Nutrients and the foods that supply them are still important. Will superfoods save us? Will junk foods kill us? Sounds dramatic, but these prescriptions oversimplify nutritional choices. Do plant-based diets guarantee happier or longer lives? Is fast food really a heart attack on a plate? Does eating only healthy foods assure a healthy diet? Or does the practice condemn us to an obsession with food and a pattern of disordered eating?
ASIDE. More than six months ago I ate my last meal from a restaurant, dine-in or take-away. Last week I ventured back into the real world of dining out. I drove through at a fast-food place and ordered a large portion of fries. I requested extra mayo. The first fry was overdone with a slight burnt flavor. The rest were bliss! I did not eat again for sixteen hours. Did my binge put my health at risk? Or did the vitamin C in the fries more than compensate for the risk associated with eating fried food?
“Random Chance is not fixable: it’s just something we have to try to understand. Unfortunately, we have managed to spectacularly misunderstand p-values for decades,” (p. 188) Zaidan saves the meat of the book for Chapter 7 of 10. He reveals the dangers of searching for statistical significance without demonstrating relevance. ‘p-hacking,’ as it’s known, can distort reality. He illustrates his point using a study done with the Registered Persons Database. This database compiles disease information on every person admitted to any Ontario hospital. Statistician Peter Austin demonstrated a serious statistical flaw when analyzing large sets of data (1). Austin partitioned his data into signs of the Zodiac. Women with the sign of Virgo were 40% more likely to experience vomiting during pregnancy. Taurians exhibited 27% more cases of diverticulitis than all other Ontario patients. And, those under the sign of Scorpio were 50% more likely to evidence anal and rectal abscesses. YIKES—I’m a Scorpio!!
Do these relationships condemn patients with specific Zodiac signs to certain medical conditions? Highly unlikely! Welcome to the world of Random Chance. Does this study condemn all scientific research to meaninglessness? No, but to overcome Random Chance researchers develop hypotheses and design experiments to test them. Traditional scientific studies do not mine huge databases for statistically significant relationships. Instead such databases tempt investigators to search for significant p-values (p<0.05). Any significant relationships are publishable even if not relevant. Austin found such relationship between the signs of the Zodiac and medical conditions. Zaidan questions all such studies linking ultra-processed foods to chronic diseases. Does eating ultra-processed food cause cancer? What about cardiovascular risk, diabetes, premature aging, reduced life expectancy, or other medical conditions? I will pursue this issue on this site later this month.
“Don’t worry so much. Ignore most news about food and health—unless it’s a safety recall or contamination notice or something like that. Health news is not designed to give you contextualized, nuanced information about your health; it’s designed to sell advertising and cookbooks. It’s the latest word, not the last word.” (p. 261) The author takes us through the seven potholes of scientific research and interpretation. When we turn to science, we expect certainty. Science delivers insight. Science does not deliver certainty. We want what we read to be the last word rather than another piece in a grand puzzle. In the middle of a pandemic, we seek fail-safe solutions and a vaccine just over the horizon. The best information available to date fails to comfort us.
Zaidan, the chemist and science communicator, rejects nutritional epidemiology. He believes these epidemiologists lack the tools to develop useful conclusions. His careful analysis reveals the dangers of misinterpreting data. I taught a graduate-level course in Food Research and the Scientific Method. In it I emphasized critical thinking skills. If teaching that course today, I would make Ingredients a case study in critical thinking. Like the author, I would explore, in-depth, p-hacking and Random Error.
Bottom line. The author concludes that ultra-processed foods do not pose a danger to public health. In his analysis, Zaidan points out that even the estimates of the health risks are misleading. Actuarial tables form the basis of the estimates. If we accept the dire predictions of nutritional epidemiologists, eating ultra-processed food increases the risk of death by 14%. This statement suggests that someone with life expectancy of 78 will die at 67. Rather, the data indicate a decrease of 1.3 years say down to 76.7. Losing 1.3 years is not nothing, particularly to a 70-year old like me! It also means that I can’t have a conversion experience at age 70 to gain back those 16 months. I needed to change my diet back in my 40s if these claims are true.
If I avoid all ultra-processed products and start gorging on superfoods today, I will not bend the curve. Such choices will have little effect on my life expectancy. My immune system deteriorates as I age. I consume Cheerios with Smucker’s sugar-free syrup for breakfast at least once a week. I eat lots of gluten-free toast and a Reese’s peanut butter cup each day. I also scoop out a small bowl of lactose-free ice cream each week topped with a spoonful of peanut butter. Delicious! Such are not my only ultra-processed sins. In my defense, I love to drink fresh orange juice and enjoy red bell pepper in my salads. Kale, not so much. The author backs off a little at the end of the book telling us to eat within our comfort zone. He destroys the argument, however, against ultra-processed food thoroughly and effectively.
Next week: How much does homemade mayo differ from store-bought mayo? by Julie Jones
Reference: (1) Austin, P.C., Mamdani, M.M., Juurlink, D.N. and Hux, J.E. 2006. Testing multiple statistical hypotheses resulted in spurious associations: a study of astrological signs and health. Journal of Clinical Epidemiology 59:964-969.