Listen on: Apple Podcasts | Spotify | Amazon Music | Podcast Index | Overcast | YouTube

Every morning I get on my bicycle, and the helmet goes on without a thought. So a few weeks ago, when a motorcyclist passed me bare-headed, my first instinct was to flag him down—because I’d just read a news story reporting that e-bike crashes had doubled, that one in five of those riders arrived at the ER with a traumatic brain injury, and that four out of five weren’t wearing helmets. The evidence for helmets is strong: a meta-analysis by Olivier and Creighton found they cut the odds of a head injury roughly in half. Shouldn’t I chase him down?

I didn’t. And the reason I didn’t is the reason for this article.

We are bombarded with health headlines. Coffee reduces dementia. Even light drinking causes cancer. Fish oil protects your heart. Some of these are important and worth acting on. Some are misleading. Almost none tell the full story. As a physician and scientist, I read medical studies every week, and over the years I’ve distilled my reaction into five questions. By the end of this piece, you’ll have a practical filter for deciding whether the next headline should change what you eat, what you swallow, or what you ask your doctor—or whether you can say “ho hum” and move on.

This is also the first of two articles. Today is about whether the evidence is actionable for you. Next time: whether the person—or the AI—explaining that evidence deserves your trust.

Question 1: Can I trust the evidence?

An interesting finding is not automatically practice-changing evidence. Picture a ladder. At the bottom sit test-tube and cell studies—a supplement that perks up mitochondria in a dish tells you about a mechanism, not about you. One rung up, animal studies suggest plausibility, not proof. Then come observational studies, which spot real-world connections that may or may not be causal. Near the top are randomized trials, which separate the treatment from the people who chose it, and meta-analyses, which are only as strong as the studies they pool.

Fish oil is the cautionary tale. People who eat more fish do show lower cardiovascular risk in observational data. But fish eaters also tend to smoke less, exercise more, and eat better—the “healthy-user” problem. When the question was handed to randomized trials, the story changed. VITAL followed 25,000 healthy adults for five years and found no significant drop in major cardiovascular events. ASCEND found no benefit in people with diabetes who had no established heart disease, and OMEMI found none in survivors of a heart attack. Eating fish and taking a fish-oil capsule, it turns out, are not the same thing.

Question 2: How big is the benefit—or harm—really?

“Cuts your risk by 97%.” “Raises it by 30%.” A percentage without a starting number is an unfinished sentence. Finish it before you decide.

The shingles vaccine is a good number. In the ZOE-50 trial of more than 15,000 adults over 50, it reduced shingles by 97%—and the absolute figures hold up: about 9 cases per 1,000 person-years without the vaccine versus 0.3 with it. That’s roughly 9 fewer cases for every 1,000 people each year, a benefit clearly worth having.

Now the other direction. Moderate drinking has been linked to about a 17% higher risk of colorectal cancer. Frightening—until you finish the sentence. Against a baseline lifetime risk near 3.9%, that raises it to roughly 4.6%: an absolute difference under one percentage point.

The same logic explains why I wear a helmet but don’t lie awake over it. Helmets cut head trauma by about 48% and severe brain injury by 60%, yet a cyclist’s odds of a serious crash on any given trip are low—one to five per million. For a motorcyclist, the baseline risk is roughly fifteen times higher, so the same relative benefit protects far more real riders. Relative risk tells you how dramatically the number moved. Absolute risk tells you whether it matters.

Question 3: Who was studied—and does it apply to me?

A study can be excellent and still have almost nothing to do with you. “Mediterranean lifestyle cuts diabetes risk by 31%” comes from PREDIMED-Plus, a strong six-year randomized trial of about 4,700 people (10.3% developed diabetes versus 12.8%). But those people were older Spanish adults, 55 to 75, carrying extra weight and metabolic syndrome—and the intervention wasn’t “add olive oil.” It was diet, calorie reduction, exercise, and counseling aimed at weight loss. A headline that names one catchy ingredient can hide a whole program.

The study describes a population; you are an experiment of one. Ask whether you share the age, the risk profile, and the actual program—not just the tagline.

Question 4: What is the downside?

Put both sides on the scale. Every real intervention carries a cost—a side effect, a bleeding risk, a surgical complication, or simply the hassle of a life overhaul like PREDIMED-Plus. The helmet is an easy call: the burden is helmet hair and a warm scalp, and the potential loss is catastrophic. Aspirin is harder—its bleeding risk is real. And some costs aren’t medical at all. Cutting alcohol has small measured benefits, but ritual, enjoyment, and time with friends are legitimate values that belong on the scale too.

Question 5: So, what should I do?

Here is where it got personal. An abnormal treadmill test raised a concern about my coronary arteries, and I went on a daily aspirin. Reasonable—until I ran my own five questions.

Can I trust the evidence? Yes: a meta-analysis of 20 trials in 100,000 people. How big is the benefit? For genuine secondary prevention—people with established disease—aspirin lowers events meaningfully. But for someone closer to primary prevention, like me, the trials show a drop from roughly 0.57% to 0.51% of events per year—about six fewer per 10,000—and no effect on mortality. That is why USPSTF and AHA/ACC guidance has pulled back. Does it apply to me, and what’s the downside? A tiny benefit against a real bleeding risk—and I like to ride my bike. I stopped the aspirin.

Then I put the helmet through the same filter. The evidence is solid, the benefit if I crash is meaningful, I ride on public roads so it applies to me (even if the scary numbers came from e-bikes), and the downside is trivial. For me, the helmet stays on.

As for the bare-headed motorcyclist—I let him go. The headline about him wasn’t wrong, just incomplete. That’s the whole point: a headline is the beginning of the story, not the end. Learning to ask these five questions may be one of the most useful health habits you ever build.

If a printable version would help, I keep a one-page action guide at drbobbyevidence.com. And I’d love to hear the headline that made you stop and wonder.