This article was first published in聽
Last semester I carried out a little experiment in my Chemistry of Food course. At the end of the last lecture, I handed out a blank sheet of paper and asked students to anonymously list the foods and beverages they had consumed the day before, being as specific about amounts as possible. I also asked them to comment on a scale of 1-10 about the impact that my 39 hours of lectures had on changing their diet and what specific change they made. I had about 400 responses and they sure made for interesting reading. Obviously, this can in no way be construed as a scientific study, I was just interested to see if anything really popped out. And it did!
It immediately became clear that even a rough estimate of calorie count, protein, fat, sugar or any other nutrient intake was impossible because of the generally vague report of quantities. 鈥淎 cup of pasta,鈥� 鈥渟erving of roasted chicken,鈥� 鈥渁 bowl of cereal,鈥� 鈥渙ne serving of strawberries,鈥� 鈥渂owl of noodles, beef slices and bok choy鈥� or 鈥減late of salad鈥� are not specific enough to allow for even a guess at nutrient intake. Memories were also quite spotty. One student listed 6 onion rings, around 10 French fries and 2 litres of oat tea as his total consumption, while some others had long, detailed lists but with sketchy guesses at amounts. Nevertheless, I was able to glean some interesting information.
Fruit and vegetable consumption appeared to be at most 2-3 servings, well short of the recommended 5-7. I was stunned by how few students listed pizza, and by the almost total absence of candies, desserts and soft drinks. It is a good bet that there was a reticence to report behaviour that may be looked on as unfavourable. Since I had put much emphasis on the need to reduce sugar consumption, had talked extensively about the problems associated with highly processed foods, and had demonized sodas, students may very well have felt guilty about reporting consumption of these items.
When it came to the changes they made in their diet, the average on a scale of 1-10 was 5, with the most frequent mention being cutting down on soft drinks, which according to the questionnaire they seem to have done. They also said that they learned to reduce processed foods and increase fruit, vegetable and fibre intake, which was not supported by the submitted data. My primitive experiment does not lend itself to any conclusion other than students鈥� diet is generally poor, reporting amounts of food consumed is challenging, and students may not want to admit consuming foods that they have learned are of poor nutritional quality.
Much of what we know about nutrition has been learned from observational studies similar to mine but far, far more sophisticated. Nevertheless, they are subject to some of the same vagaries, namely difficulty in recalling exactly what has been eaten, inability to estimate quantities correctly, and reporting what they think they should have eaten instead of what they actually ate. Food frequency questionnaires, the basis of most epidemiological studies, are in reality a reflection of the subjects鈥� perception of what they consumed, not what they actually ate. We are stuck with these hitches because interventional studies involving diet are extremely difficult to carry out.
For example, if we wished to determine whether the daily consumption of a cup of blueberries has an impact of health, we would have to randomly divide volunteers into two groups of significant size with both groups following identical diets except for the consumption of blueberries. To ensure that the diets are identical, meals would have to be provided and adherence monitored. To yield meaningful results, the study would have to last at least a decade or more. This is economically, logistically and administratively almost impossible. So, we are left to rely on studies in which researchers observe and collect data on groups without actively intervening or manipulating any variables.
Perhaps the most ambitious such observational study on the relationship between nutrition and health is the French study that began in 2009. By 2021 approximately 171,000 people had enrolled, making it the world鈥檚 largest ongoing nutrition study. Publications based on the collected data began to emerge in 2017 with the first paper reporting an inverse association between organic food consumption and Type 2 diabetes. A subsequent publication reported that a higher frequency of organic food consumption was associated with a reduced risk of cancer. At this point it is important to emphasize the major shortcoming of observational studies, namely that 鈥渁ssociations cannot prove a cause-and-effect relationship.鈥� For example, it is possible, in fact likely, that people who buy organic food are more health conscious, are more physically active, and follow a better diet.
By 2025, the wealth of data collected had fuelled a number of papers in diverse journals with some noteworthy observations. Support was found for a beneficial role of higher intake of plant-based products along with lower intakes of animal products in the prevention of cancer. Ultra-processed food consumption with an increased risk of cardiovascular disease and earlier mortality was also reported. Another paper associated sugary drink consumption with an increased the risk of cancer. But subjects who replaced sugar with artificial sweeteners were not better off. Sweeteners, especially aspartame and acesulfame potassium, were associated not only with increased cancer risk but also with increased risk of cardiovascular disease. Even the relationship between dining schedules and disease has been examined in 狈耻迟谤颈狈别迟-厂补苍迟茅. A pattern of eating breakfast before 8 a.m. and an early supper followed by fasting till next morning was linked with a lower risk of cardiovascular disease.
Researchers keep digging into the data and are searching for whatever can be extracted, including possible adverse effects of certain food additives. A mixture of modified starches, pectin, guar gum, carrageenan, polyphosphates, potassium sorbates, curcumin and xanthan gum was found to be associated with Type 2 diabetes. I don鈥檛 know to what extent the data had to be tortured to yield this result, or what one would do with this information.
While the interpretation of data by 狈耻迟谤颈狈别迟-厂补苍迟茅 scientists has received criticism, the large number of subjects involved allow for some broad conclusions. Adhering to a mostly plant-based diet while cutting down on ultra-processed foods and curbing both sugar and artificial sweetener intake is likely to yield health benefits. That may now sound like old hat, but remember that the hat was originally produced by observational studies such as 狈耻迟谤颈狈别迟-厂补苍迟茅.
