The trouble with French fries is not the oil

We struggle with interpreting, accepting and/or being mislead by "studies"

Secretary of Health and Human Services Robert F. Kennedy Jr. is right when he says that chronic disease is on the rise in America and that our food system is at least partly to blame. Where he and his “Make America Healthy Again” movement err is in relying on flawed evidence to target particular foods.

Let’s take seed oils. Mr. Kennedy has claimed that oils made from seeds — sunflower, safflower, canola — have “poisoned" Americans, and are “one of the driving causes” of the obesity epidemic.

The fear over seed oils stems from the fact that they tend to be high in one type of unsaturated fat, omega-6 fatty acids, whereas olive oils have more omega-3 fatty acids. Studies have found that people who consume a higher ratio of omega 6 to omega-3 fatty acids in their diet tend to have worse health.

The problem is that most of those studies are poorly designed to show how omega-6 fatty acids, or seed oils that contain them, are actually causing the negative health effects.

Seed oils are prevalent in processed and fast foods. That’s because they are cheap and also have a high smoke point, which is good for frying. There’s evidence that many ultraprocessed foods harm our health, but is the problem the seed oils, or the foods they are used in?

A meal from Taco Bell is more likely to contain seed oils than a salad and chicken breast cooked at home with olive oil, but the oils probably aren’t the reason the former is less healthy. People who eat a lot of ultraprocessed foods may differ in other ways that affect health, like their education levels, exercise habits and whether they smoke.

Research papers commonly address this concern by trying to adjust for differences across individuals, but they simply can’t control for all the factors that are associated with different dietary choices.

It’s not just seed oils. Studies that show only patterns, not cause and effect, are pervasive in the field of nutrition research.

Those headlines linking red meat to cancer or coffee to longevity are based on studies that find connections between foods and health but that cannot determine if one causes the other.

When we fixate on individual foods or ingredients, we miss the bigger picture — implying that chronic disease would be fixed by, say, removing food dyes from Skittles.

Often, with better data, the conclusions of nutrition studies based on observational findings turn out to be wrong. For decades, people thought fat was the enemy and dietary guidelines encouraged people to eat less of it, and more carbohydrates.

Then, a landmark study in 2006 showed that women randomly assigned to follow a low-fat diet were no less likely to suffer from heart disease or stroke than people who ate more fat.

It is not that we know nothing about nutrition. A large randomized trial has shown benefits of a Mediterranean style-diet, and evidence from high-quality nutrition studies has shown a diet high in ultraprocessed foods results in eating more, leading to weight gain.

The general advice to eat mostly whole foods and not too much is reasonable.

Where the evidence falls apart is when it turns to specific foods.