Today we will cover a list of 13 sleep factors that you can modify to get more bang out of your time spent sleeping.

You probably spend about 25% to 37% of every 24 hours sleeping. (That works out to 6 to 9 hours). If we are all spending that much of our time doing something, we might as well get the most benefit out of it as we can.

This article provides an in-depth guide on the core principles and practices that you should be aware of. This is worth your time if you want to be someone who takes their sleep seriously and who wants to gather as many health benefits as possible from those hours each night.

My own sleep has gotten much better in the past year or so (not-so-coincidentally at the same time as our daughter has been consistently sleeping through the night). And yet, though I get ~7-8 hours most every night, I know that there is room for me to improve.

If I’m being totally honest, I chose the topic for todays’ letter with the goal of transforming myself from some who sleeps “pretty well” to “amazing.”

That’s a side benefit of doing this work - I find that I often end up learning (or re-learning) things that I can apply directly to my own health or that of my family.

As we will discuss, your sleep quality is associated with your health in a very serious way. Sleep measurably affects your immune system, deciding whether you get sick and even deciding how well your vaccines work. In long-term studies, sleep quality is associated with dementia risk - - 25 years down the road. In studies as short as 1 week in duration, changes in sleep cause drastic differences in blood sugar control and insulin responses (including in my wife, who did a small n=1 study on herself quite by accident back in fellowship, as we will discuss).

Note: This article has a “quick takeaways” summary at the top of the extended section, but the full letter is the length of a full book chapter. If you are reading in your email browser, it will almost surely be truncated. I suggest that you either click on the title at the top of this letter (which will open the article in your web browser, where you can read the full thing), OR you could read it directly in the substack app. Or, you could continue to read here in your email until you get to the “…” symbol that you can click on to expand the full letter. Just don’t be surprised if it appears to end abruptly.

Let’s start with a look at what happens to your immune system when you don’t get enough sleep.

Researchers squirted virus up the nose of healthy volunteers who slept different amounts

The quality of your sleep influences your odds of whether you get sick, how sick you get, and – to some extent – how well your vaccines work. Getting high quality sleep is a measure of direct biological defense.

If you have not run across “challenge studies” before, they sound like they were dreamt up by evil henchmen, perhaps who really turned out to be a stack of virions in a trenchcoat.

Challenge studies recruit healthy adult volunteers, squirt some virus up their nose, and then… see what happens. They are a common mode for the early phase testing of some vaccines and therapeutics, and also for answering fundamental biologic questions (like the impact of sleep on viral defense).

In one such challenge study, researchers squired rhinovirus into the nose of volunteers after measuring their sleep duration and sleep continuity over 7 consecutive days. They calculated that participants who slept <5h per night had significantly greater odds of developing a cold compared to people sleeping >7 hours per night:

The researchers found that this effect appeared independent of participant immune status (antibody titers), demographics, season of the year, BMI, psychological variables, and measured health practices. Below, we will cover other research that suggests that the quality of your sleep in the days around vaccination can also affect the level of protective benefit you get from a vaccine.

Notably, this study had more than 100 people, but the effect size estimates were very wide. The 95% confidence interval for the odds ratio of <5 h sleep compared to >7h sleep was 1.08 to 18.69, encompassing “a trivial difference” to “a massive difference.” So, the exact size estimate on this effect is rough, and interindividual variation is huge.

How much money would it take to get you to sign up to be in such a challenge study?

I think we all intrinsically understand the effect of sleep on our well-being and our health. What is less obvious is how quickly it can affect some of your health metrics, to what extent, and which common “fixes” actually do nothing to improve your sleep.

The good news is that sleep is one of the rare high-yield activities that you can improve while costing very little in terms of $ and time, since you are going to be sleeping anyway. In fact, some of the sleep “interventions” that cost the most money actually have some of the weakest supporting evidence. In any event, based on my reading of the sleep literature, I expect that many people are leaving sleep quality on the table for entirely fixable reasons.

That is the purpose of today’s letter – to share those findings and help you get the most you can out of your Z’s.

The full guide is below

As you know, this letter has no advertisements, no sponsor, and nothing to sell you here other than my time spent interpreting the evidence.

Here’s why I do this as a paywall and not an ad: I answer to you, full stop. On a topic where half the internet is trying to get you to click their affiliate links to purchase their magical sleep-optimization supplements, I’d rather you pay me for my unfiltered analysis than have a supplement company pay me a percentage of sales for something with questionable supporting evidence.

Extended health articles are available to my paid subscribers in return for their support. Here’s everything in today’s letter, with every study linked so you can check my work:

• The single highest-yield change to make - and it doesn’t require sleeping more. A ~60,000-person study found it predicted death on par with how long you sleep.

• What are the stakes?

  • Immune system

  • Brain health / Dementia

  • Metabolic Health

  • Mortality

• How much sleep should you target? What to do about morning and evening light? What about sleep temperature settings?

• The gold standard for insomnia (which is not a pill)

• Popular claims that do NOT work

• Red flags that should send you to your doctor

A paid subscription also opens the full archive: every deep-dive I’ve written, and everything still to come. If you’re ready to dig back, a few popular prior articles are:

• The Truth: Is Alzheimer’s Preventable?

• Creatine: What Works and What’s Hype

• The #1 Predictor of How Long You’ll Live

• How to Reduce Your Visceral Fat

Subscribe now

Eight dollars a month (that’s the average via the $96 annual plan, which works out to less than a monthly bottle of the melatonin gummies I’m about to tell you not to buy), and you’ll walk away knowing exactly what’s worth your effort vs. what’s a marketing trick.

If it’s not useful, unsubscribe any time in one click.

Let’s get into it, starting with a summary of the 13 different sleep factors, the targets to set them to, and a one-liner on how they work. If you don’t have time to read the full letter, at least read this table of critical sleep factors:

Read more

Here’s a headline that went around recently, and I’d bet it made you do a double take:

The story underneath said healthy older adults who took sugar pills for three weeks got “real boosts in memory, movement, and stress,” even the ones who were told, flat out, that the pills were placebos!

Well that sounds like something worth tuning in for.

A fake pill that sharpens an aging memory would be a big deal. A reader sent it to me and asked: is this true?

I pulled up the study. It turns out to be a small, modest paper that was tortured via press release into making claims that it never asserted. That insiduous transfiguration, from what the researchers wrote into what the headline declared, is a common theme in health misinformation, so perhaps this analysis can provide some broader lessons to consider when you encounter future headlines.

The study had three groups of participants

1) Received a placebo pill, but were told it was a multivitamin

2) Received a placebo pill, and were told it was a placebo, but also a motivational speech about how placebos can really affect the body.

3) Received no pill and no inspirational speech about the placebo effect.

Everybody was given a series of tests at the beginning of study, and then again 3 weeks later. They examined changes in those pre-/post-results

First note: a useful trick for reading any “taking X improved memory by Y%!” story. It comes down to one question with two options that sound the same but are not:

• did the treated group improve

• did the treated group improve more than the control group?

Let’s examine what really happened in this “placebo improves memory” study.

Below, for paid subscribers:

• What held up and what did NOT: We examine what the researchers really measured, and how that differs from what the headlines claimed.

• A consideration on what a study like this can and can’t add to the science. Is there even a point?

• Three questions I consider on any study like this, so you can give the next miracle-pill headline the same level of scrutiny.

Paid subscribers get the full, sourced deep dive behind the health headlines, twice a week, for less than a dollar a letter. I take no ads and no sponsorships, and there’s nothing in your cart at the end. That independence is the point: it’s the reason I can tell you when something popular is being oversold.

The premium newsletter is available at $12/month or $8/month (averaged out on the annual plan). You’ll leave every issue feeling more informed about your health.

Now, let’s get into it.

Read more

Spend a minute looking at the below figure, which depicts that very same data.

Upon a close inspection, you will find that there is something very different at the very bottom of the household income percentile curve, and that going from the bottom 1% of income to even just the bottom 3% of income is associated with a substantial improvement in longevity. But even if you start from that 3% mark and compare against the wealthy 99% values over on the right, it is clear that there is a “dose response” effect of household income on longevity for both men and women.

Of course, the grim reaper does not decide when to come a knocking after a careful review of your tax returns, brokerage account statements, or avocado toast receipts.

The easy interpretation is that rich people can buy things you simply cannot.

Some of that is true. For example, the wealthy typically do less physically dangerous work (they are less likely to work as a logger, commercial fisher, or roofer), they experience less stress from chronic financial panic, and they end up in better-staffed hospitals when something eventually goes wrong. The playing field is not level at any stage of the process, at least not in the United States.

But when you itemize many of the longevity-promoting factors that money can actually buy you, the tradeoff often works out to one of time, money, and convenience, and only rarely is it a hard barrier to access.

So today we will do a thought experiment:

Budget: unlimited.

Goal: Decrease risk for cardiac disease and cancer, decrease risk for dementia, and maintain muscle and bone health well into your 80s+.

I have identified a few categories of expenses that a wealthy family could make. For each, I will point out the biological health factor that the purchase is really affecting, and then briefly discuss how to achieve the same health effect for much less money (at the cost of your time).

To be very clear up front, I am a scientist and none of this is medical advice for your specific case, so run the details past your own doctor.

There is no rich-person biology

To this day, there are some things that money simply cannot buy… like a molecular pathway your body doesn’t already have. Jeff Bezos ages on the same cellular machinery you and I do.

No fortune has yet bought anyone an exemption from the woes and perils of being a mammal.

What money certainly can buy is friction removal. Willpower has very little to do with this discussion. What the money builds is a life where the healthy choice is the default, already scheduled, already sitting in the fridge when you walk in hungry at 6pm with four minutes on the clock before your child needs some help with their watercoloring activity.

Most of the below factors come down to buying back time and buying easy adherence. The biology is relatively straightforward and is the same for everyone. What money buys is the convenience of making it easier to do. And at the end of the day, that’s what counts, because it doesn’t matter what you know about i) what you should be eating, ii) how you should be training and sleeping, iii) what your blood pressure or cholesterol should be, or iv) the effects of your home air quality / PM2.5 - - what counts is what happens in the real world and what your body experiences on a daily basis.

#1 - Hire a personal chef to make it easy to get 30+ grams of fiber a day and fewer ultraprocessed meals

Your private chef’s technical scope is to shop for ingredients and come into your house once per week to prepare 20 meals (10 each of two recipes).

The Cost: about $450-600 a week depending on where you live, plus groceries, so on the order of $25k-30k per year.

The fantasy interpretation of this purchase would be something akin to a longevity potion prepared with a mystical blend of elusive botanicals. In reality, the chef is actually plating more fiber, more plants, and resulting in a kitchen with less ultra-processed food to raid on a whim. Imagine opening your fridge door and finding neatly packed and labeled glass Tupperware, and the only “cooking” you have to do is choose which one of these to microwave:

• Mediterranean Chickpea & Veggie Bowl with Herbed Couscous and Mixed Nuts

• One Pan Mediterranean Fish with Roasted Zucchini and Orzo

• Chicken and Lentil Curry with Mashed White Beans, Farro, and Roasted Carrots

This is all about reducing friction at the time you are making a decision, and making the pro-health option just about as easy as the packaged & convenient snack option.

The Benefits:

Dietary Fiber - The fiber number that your chef will help you hit every day is one of the more startlingly underappreciated stats in nutrition. (Unless you have read a few of my prior posts or know a registered dietitian, in which case you will never hear the end of the pro-fiber discourse).

In a large Lancet review, the people eating the most dietary fiber were 15 to 30 percent less likely to die early, from any cause and from heart disease, than those eating the least fiber. In raw counts, that’s about 13 fewer deaths for every 1,000 people, when comparing the highest fiber intakes to the lowest (Reynolds et al., Lancet 2019).

Your rough daily target should be to get at least 14g fiber per 1,000 calories, which works out to a daily target of ~25-35g+. Most Americans get about 15g, roughly half the target. What the chef really does is make hitting your daily 30+ grams effortless and delicious, so you actually hit it most days for decades, which is the timescale that counts.

Ultra-processed foods - The flip side of the chef’s cooking is what it crowds out. Higher intake of ultra-processed food tracks with a modestly higher risk of a long list of bad outcomes. A 2024 review pooling 45 analyses across roughly 10 million people linked it to higher rates of early death (a bit over 20 percent higher, comparing the most ultraprocessed intake to the least), heart disease, and depression, though the authors themselves rated most of that evidence low quality (Lane et al., BMJ 2024). A more concrete signal comes from the one tightly controlled feeding trial that I could find, where the NIH housed 20 people in a metabolic ward and provided them with ultra-processed food options matched for levels of sugar and fat and fiber against an unprocessed one. Each person spent 2 weeks with ultraprocessed options and with unprocessed options. On the ultra-processed weeks people ate an average of about 500 more calories a day. Over the 2 weeks, they gained an average of 2 pounds. And when people switched to the unprocessed diet for 2 weeks, they lost an average of 2 pounds (Hall et al., Cell Metabolism 2019). A small and short-duration study, but it provides evidence that the food itself drives some of the overeating, independent of who’s eating it.

Of course “Ultra-processed” is a rather crude designation that lumps soda and hot dogs in together with whole-grain bread and canned beans. When researchers pull it apart, the harm concentrates in a couple of usual suspects, sugar-sweetened drinks and processed meats, while ultra-processed breads and cereals come out roughly neutral (Cordova et al., Lancet Regional Health Europe 2023). Careful with that last part, though: those are subgroup findings from observational data, the slice where confounding bites hardest, so read them as “soda and deli meat are the clearest offenders,” not as proof that a packaged whole-grain loaf is health food. The practical takeaway is to cut the sugary drinks and processed meat first, and stop losing sleep over whether the bread came in a wrapper.

The “free” alternative? You already know how to save yourself the $30,000 chef, annoyingly: it comes at the cost of a couple hours of shopping and a few hours of batch cooking. Search for recipes that naturally incorporate lentils or beans and roast a couple sheet pan of your favorite vegetables, and portion them into grab-and-go containers. Make a jar of overnight oats or chia pudding for the week’s breakfasts in five or ten minutes. Keep cut fruit and plain yogurt at eye level in the fridge, and keep the things that you don’t want your future self to eat either out of the house or somehwere with a little more friction. Friction works both ways: you lower it on the pro-health meal and snack that’s already cooked or prepared, and you raise it on the snack you never brought home.

The next purchases are standing appointments with your trainer, a concierge doctor and an executive physical, a premium leafy, walkable neighborhood, and staff to buy back your time to make it all work in your schedule.

The playing field is NOT level (see the figure above). But each of these purchases is buying a specific, dull factor, and each has a free route below that trades time for money.

In today’s extended letter:

• The personal trainer, who is really buying you four key things: bone density, muscle mass, aerobic fitness, and the scheduled calendar events to make sure you actually show up. Also, thoughts on adherence without the $20k price tag.

• The concierge doctor, and how to rig it yourself with the exact numbers to walk into a normal 15-minute primary care visit asking for.

• The right house, which is protecting your brain more than you’d think, and how to copy some of its active ingredients.

• The household staff, and where they fit in with the other factors.

• The Shopping List: in each category, what you can buy to engineer your own environment in favor of your health for a few $100’s, rather than a few $100k’s.

Subscribe now

I run no ads, take no sponsorships, and use no affiliate links. That independence is the whole point. I would much rather you pay me directly for my time spent reading and translating the evidence, rather than to take money as a cut of sales from some product I recommend you buy. For example, you will notice that, below, I recommend you consider a HEPA filter for your home air quality. Though I mention specific brands and models I have bought, I have no financial relationships with those companies. When we talk about blood pressure monitors, I point you toward a list of validated machines, rather than choosing a device that would give me the biggest affiliate kick-back. When I wrote on the benefits of creatine monohydrate, I didn’t give you a link for any specific product. I think that’s the only way you can take any of this seriously.

A paid subscription also opens the archive and gives access to every deep-dive yet to come. If you’re ready to check out some prior letters, two popular ones are: i) The Truth: Is Alzheimer’s Preventable?, and ii) Creatine: What Works and What’s Hype.

The monthly option is $12/month and the $96 annual option works out to $8/month (which is about $1 per deep dive letter).

Onward, let’s start with the overview in grid format for each purchase, the health factors it is really modifying, and the outcomes that it affects.

Read more

-Morgan McSweeney, PhD (@dr.noc)

I recently wrote about the twelve minutes you get with your doctor, and how easy it is to walk out feeling like you spent them badly.

Maybe “badly” is too strong of a word. But it sure is easy to feel like your precious few minutes speaking with your doctor were not very efficient.

One comment on that prior letter that caught my eye was from a reader named Chris, who mentioned he is a practicing primary care physician:

I am a practicing PCP. I am immediately trying to figure out the best way to share this post with my patients.

Today’s letter expands on that prior one to help you take advantage of your next visit.

My wife is a physician and many of our close friends are physicians, so I put the following question to a handful of them:

Interestingly, not a single answer came back about medical knowledge - Nobody texted me back to say they wished their patients had read more studies in advance of their visit. No one pined for a cracked textbook.

But several of them did wish their patients would address the scary item first.

Here are some of the things they told me.

Your doctor is not grading you on how much you know, or how much you care about your health

Your visit is not a test of how informed or disciplined you are, and nobody is secretly giving you a grade. Several physicians underlined the importance of this.

Your doctor is a person trying to solve a problem with the information you give them, in about twelve minutes, and the plan they build can only be as good as that information.

They need accurate info.

Trust me, I’ve heard and seen it all, and I’m immune to even being a little bit judgemental. I just want accurate info so I can help them as best I can.

Some of the other suggestions below will feel a little embarrassing in the moment, but all of them are powerful precisely because you have dropped the subconscious performance of being the “informed patient who cares about their health.” When you do that, you are ready to start providing the information they need to make a well-informed decsion. And the sneaky thing here is that the performance of being an informed patient can happen subconsciously unless you make a direct mental effort to combat it in the minutes before your doctor comes in.

Last year, I personally had to switch primary care providers because my former one moved to an office that was 45 minutes further away.

(I had liked him quite a lot, but not enough for a 2h round trip).

When I next met with my replacement doctor, she asked me what I do for work, with a slightly suspicious aura that I have developed an innate ability to recognize over the years.

When I said that I work in biotech and that I do science communication, she said “I thought so! I follow you on Instagram!” Then, she never mentioned it again and went straight back to doctorly business. But, even though everything that followed was 100% professional and well-handled, I decided to switch doctors again to someone who did not know me as “a guy who talks about health and science.”

Not because of anything that my prior doctor did wrong, but because I knew that it would bias me - - subconsiously or otherwise — toward feeling like I should adopt the role of “the informed patient who cares about their health,” and I know how dangerous that can be.

So I switched. If you do not feel like you can be 100% transparent with your own doctor, I would encourage you to switch to another provider that you feel more comfortable sharing the nitty gritty accurate details with. Not everyone is a good fit with everyone else.

You have mere seconds before the agenda goes in its own direction

In a 2019 study of primary care visits, doctors invited the patient to lay out their full set of concerns only 36% of the time, and when a patient did start talking, the clinician redirected them after a median of eleven seconds. An earlier and now-famous study found patients got to finish their opening statement in fewer than one in four visits.

This is not really a knock on doctors. Eleven seconds is perhaps what we should expect to happen when you ask a busy, well-meaning person to fit six months or a year of your health counsel into twelve minutes: they start steering early because they have to, and because over the years they have developed a pattern recognition system that gives them an innate sense of where the conversation is headed. But it tells you exactly what to do.

Walk in and put the entire list on the table in the first dozen seconds, ranked, before anything drifts. Say how many items you have and open with the scary one:

“I have three questions today. Number is the big one, the chest tightness when I walk uphill. Number two is about my cholesterol medication, and number three is about my knee.”

If you start off with: “I have three questions,” they have made a mental note that you’ve got two more coming, even if the first question leads directly into discussion.

When I read that out loud it took me 9.5 seconds, and yet it makes sure that the scope is laid out for everyone to see.

You should write this opening statement down on a physical piece of paper or notebook and bring it with you into the room. A note on your phone is not sufficient.

Your memory, keen as it may be, is woefully vulnerable to distractions of the moment.

A physical notebook has no distractions and, more importantly, serves as a physical signal that you have something specific you want to cover. No matter how distracted you get in conversation about point #1, you’ll find yourself still holding the notebook that has #2 and #3 written right there.

Several of the doctors I asked described the doorknob effect: where the thing the patient actually came in for shows up in the last thirty seconds, when their hand is already on the doorknob. “Oh, by the way, I’ve been getting this chest pain.”

By then, the visit is basically over, so the concern that needed the most time gets the least. That’s why you put the scary one first, while there is still time to do something with it.

There is even research on the exact wording your doctor uses to get there.

Clinicians are taught that asking “Is there something else you’d like to address today?” surfaces far more of what a patient actually came in for than “Is there anything else?” and in one trial that slight wording change eliminated 78% of patients’ unmet concerns.

But you don’t have to wait and hope your doctor phrases the question the right way. Bring the “something else” in with you, scribbled on something physical, and say it at the beginning of the visit.

Beware the illusion of explanatory depth

Last time I wrote about the illusion of explanatory depth, where being familiar with something feels exactly like understanding it, right up until you try to explain it and find nothing there.

Psychologists Leonid Rozenblit and Frank Keil ran a now-famous set of studies. They’d ask people to rate how well they understood something ordinary, a zipper, a flush toilet, how a helicopter flies, and people would rate their understanding fairly high. Then, they’d ask them to actually explain it, step by step.

Confidence levels plummeted.

Over and over, people discovered that what had felt like understanding was really just familiarity. They’d seen the thing a thousand times, so they assumed they understood it. When put to the test, they did not.

You probably feel like you have a general understanding of how a zipper works. But take a moment to be honest, how well do you think you could explain what happens inside the mystical magical Box of the Zipper? How does it hold together so nicely once zipped?

We fall under this illusion of explanatory depth constantly, and we cannot feel it… because that’s the whole nature of it.

In practice, this can happen:

Your doctor says “your A1c is creeping up and we’ll want to keep an eye on it”, and then of course you nod because it sounds reasonable and from the context of that statement, you intuitively understand that higher A1c is not a good thing and that you should measure it again later to make sure it didn’t go up more.

But then, somewhere between the check-out desk and the parking lot, you realize that, if pushed on it, you couldn’t say what your A1c reading actually was, how it’s trending, what it’s supposed to be, or what “keep an eye on it” means you should actually do (or not do?) over the next few months.

The solution to the Illusion of Explanatory Depth

The good news is that there is a specific defense agains the illusion of explanatory depth called teach-back: after explaining something, a clinician asks you to repeat it back in your own words, because that is the fastest way to probe for an absence of explanatory depth. So, before you leave or move on to the next item on your list, say the plan back: “So I’m going to double the morning dose, and if my home readings are still over 140 after two weeks, I call you.”

If they say yes, write it down as a concise note (on that physical notebook you brought). If you can’t say it back, you didn’t really understand it, and that exact minute is the best time and place on earth to discover that.

The most useful sentence you can say in a doctor’s office is, “Can I say that back to you, to make sure I’ve got it right?” No one will be annoyed by that question.

Turn “keep an eye on it” into a number, a threshold, and a date

Vague guidance is an ancient enemy of successful follow-through, just as much as my dog is the enemy of the squirrels that try to bury nuts in our backyard.

Franklin.

“We should watch that” is about as vague as it gets. For any number your doctor flags, ask three things:

• what are we aiming for?

• what reading would make you want to hear from me sooner?

• when should I get it rechecked?

That means that a statment like

“Your A1c came back at 7.4, let’s work on that”

turns into a much more helpful:

“let’s aim to get it under 7, call me if you find you’re getting really thirsty, peeing a lot, or losing weight without trying, and let’s recheck in three or four months.”

The latter gives you something to really work with.

Tell the truth, especially when it comes to the number you’d rather round up

About half of people don’t take their long-term medications exactly as they were prescribed. Your doctor knows this is the case. They know that “take once daily with a meal” can easily turn into “I take it most days, when I remember.”

But it sure does help improve their decision making if they have a 100% accurate understanding of how you are actually taking your medicines, before they render decisions on dose adjustments or medication swaps.

Tell them the real number, because they are about to make a real decision with it.

If you say you take your blood pressure pill every day, when the truth is closer to four days a week, and your pressure reads high, then the logical next move might be a second drug, or a higher dose of the first, when in reality the problem might really just be those three missed days.

The accurate answer, even when it is “honestly, maybe half the time,” gets you a better plan and sometimes fewer pills.

You must resist the default (and very reasonable) temptation to sugarcoat reality when talking to your doctor. This applies to questions about drinks, exercise, vaping, sexual activity, supplements, all of it. Your doctor is working together a complex recommendation on a dose or a change, and a flattering input will lead them on a hunt for a wild goose in a room full of rabbits.

While you’re at it, bring the actual list of everything you take. Either bring the bottles in a bag, or a photo of every label on your phone, including the supplements and the over-the-counter things you don’t really quite think of as medications.

Subscribe now

Leave with three answers written down

There is a health-literacy tool called Ask Me 3.

Ask Me 3 has three questions:

1. What is my main problem?

2. What do I need to do?

3. Why is it important for me to do this?

Don’t leave the room until you can answer all three, and then write the answers down before you step out of the door.

What to write in your notebook before you go to the office

Before you go to the doctor, write down:

• Your top three concerns, ranked, with the biggest one first.

• The Ask Me 3 questions.

• Every medication and supplement you take, and how often you actually take it.

Before you leave the exam room

• “Repeat Back” the most important points to your doctor

• Write down your Ask Me 3 answers.

Before you leave the office

• At the checkout desk, get the date of your next visit, and put that date in your calendar right away.

This will take a few minutes of extra time and it will cost you a little bit of pride, but it can help turn your twelve minutes from a subconscious theatrical performance into a productive problem solving session.

A few additional bonus pediatrics insights:

This Saturday morning, expect a deep-dive article with a fun thought experiment: if money was no object, what could you buy to optimize your health? And, for each purchase category, what is the DIY (free, or at least much less expensive) way to achieve similar health benefits?

Best wishes,

Morgan

Morgan McSweeney, PhD (@dr.noc)

-Morgan McSweeney, PhD (@dr.noc)

Yesterday afternoon, I was sitting in the kitchen with my daughter, looking out the window as a massive summer thunderstorm was about to roll in. The wind was already here ahead of the rain, gusting through the trees and bushes, to which she exclaimed with a giggle, “They’re dancing!!!”

Followed soon by “…Why?” (her favorite question at the moment).

Hm, why were the trees dancing in the wind?

“Well, the wind is blowing air all around the trees, making their leaves flap around and moving the branches.” To which she responded… “Why?”

After a few rounds of answers and “Why’s?” I decided it was time to reach for the big guns: Thigmomorphogenesis. (“Thig-Ma-Morpho-Genesis”).

“When the trees get blown around by the wind, they say ‘ooh, I better get stronger for the next storm!’ and so they make their branches and trunk stronger. Wind helps them grow strong.”

Thigomomorphogenesis describes the surprisng fact that plants grow differently in the presence/absence of mechanical stimulation (like touch or wind). In an experimental biodome in Arizona called Biosphere 2, researchers found that trees grown in the isolated environment grew tall very quickly, but that some collapsed unexpectedly before reaching maturity. Although the biosphere provided the ideal levels of stuff like water, sunlight, CO2, etc., it could not recapitulate the mechanical stress from wind (even with the HVAC systems), so the trees grew fast but brittle, eventually snapping.

That’s kind of the same kind of thing that can happen to our bones and to our muscles.

Fifty-four million American adults are slowly losing bone density, and almost none of them can feel it happening. That’s about half of everyone over 50. There’s no ache and no warning, right up until something goes snap.

In fact, a woman’s lifetime risk of breaking a hip is about equal to her combined risk of breast, uterine, and ovarian cancer (Bone Health & Osteoporosis Foundation). If you’re under 65, there’s a good chance no one has ever measured your bone density, and an even better chance you couldn’t tell me what the number was, even if you have previously had it measured.

First, a word about conflicts of interest. I do not sell any bone-health supplements, no hormone balancing products, nobody sponsors this newsletter, and there are no affiliate products for you to purchase. This is entirely reader-funded. I have spent a lot of time thinking about the best model for a health-research letter like this one, and I have come to think that is the least conflicted and most transparent way to do this work. The tradeoff is that I will eventually ask you to pay for the extended version, but that my writing will never be influenced by an advertiser or a supplement company.

There is one critical fact to appreciate upfront. It is easy to think of our bones as being like stones or rocks, but that is actually quite wrong. In my mind, bones are more like muscles than stones. Bone is living tissue that changes all the time, and it responds to the demands you place on it. It responds to load. The same body whose default is a slow trend toward dismantling your skeleton can be told to rebuild it, at any age, including yours.

Online, you are at risk of coming across a different message, along the lines of:

“You’re fragile now, take your calcium pill, and, whatever you do, don’t lift anything heavy!”

That message is wrong, and I’ll show you some clinical trial data that says the opposite.

Note: this letter goes into depth on bone loss prevention / reversal. The full letter is now the length of a book chapter, which means this it may be truncated if you are reading in your email client. In that case, you can either click the article title at the top to read the full thing in your browser, click the option to expand the email at the bottom (“… Load more” or some similar message), or read in the substack app. I personally actually like reading via the app.

Who this is for (yes, you too)

If you’re a woman over 50, this letter is squarely for you, and you probably already know it.

But if you’re not in that group, I’d encourage you to not drift off, because you have more skin (bones?) in the game than you think.

If you’re a younger woman, you’re at the stage where this is a winnable fight. You build about 85 to 90% of your adult bone by age 18, and you reach your peak somewhere around age 30 (BHOF; Weaver et al., 2016). After that, the game shifts from building bone to keeping it. So the single highest-leverage thing you can do is bank a big balance early and then slow the rate of loss, with regular load and protein and not smoking, because you’ll want some buffer to carry you through the accelerated loss of menopause a couple of decades from now. Picture the slope of your bone density over time. Over a long horizon, even small changes in the rate of yearly loss lead to big differences down the road.

That’s the whole reason I want you reading this today. I’m urging a small set of weekly behaviors that lower your risk of future bone trouble, and they work even more powerfully the sooner you start.

If you’re a man, this remains relevant for you, though admittedly later and with less warning, because you (and I) don’t have a sharp menopause cliff for bone loss. And yet roughly one in five men over 50 will break a bone from osteoporosis, men are somewhere around a quarter of all hip fractures, and men are two to three times more likely than women to die in the year after a hip fracture (Endocrine Society; IOF). So it still counts. Watch out especially if you’re on long-term steroids or hormone therapy for prostate cancer, both of which can run the risk of accelerating bone loss.

And in any event, you almost certainly love someone who is on this road already, or who one day will be. The strength and balance work we will discuss is for every aging human who would like to choose when they end up on the floor, and with what velocity they approach it.

However, before we get too excited about barbells, we should take a moment to acknowledge that not everyone is capable of doing the exact types of physical activity that we will cover below. Some medical conditions make it such that even getting out of bed is impossible or a major barrier, let alone loading up in the squat rack. That could be from neurological conditions, chronic pain, heart or lung disease, the long recovery after a stroke or serious fall, or something else. The underlying principles we will discuss still hold true, and can be adapted to your circumstance. Bone and muscle respond to whatever load you can give them, and “whatever you can” is a moving target. Light resistance bands from a chair or bed, standing up and sitting back down a few times, pressing your hands into a wall, a few minutes of movement with someone there to steady you: for a body that is deconditioned, that is training. It loads a system that has been sitting unloaded, and it can produce meaningful gains from a very low starting point. These are the exact type of situations a physical therapist trains for, and you should ask for a referral from your doc if this sounds similar to your circumstances.

You can’t feel the density of your bones

Osteoporosis is one of many so-called silent diseases with no symptoms. Much as your muscles won’t ache as they slowly shrink over time, your bones won’t ache while they thin. Unfortunately, that means you’ll feel totally fine until something eventually snaps.

You’ll feel exactly the same at a “T-score” of -1 as you do at -3.

That T-score is the number used to grade bone density. There are age-dependent norms, but to give you a loose idea, around -1 and above is normal. Between -1 and -2.5 is osteopenia, the thinning zone. At -2.5 or below, it’s considered osteoporosis. And if you’ve already broken a bone from a minor fall, you’re treated as having osteoporosis regardless of what the number says.

Around the time of menopause, the rate of bone thinning speeds up greatly. In the five to seven years after her final period, a woman can lose up to 20% of her bone density (BHOF). That’s a lot. And most women have no symptoms and no sign that it’s going on.

Then, one day, you catch a foot on a rug that someone else shifted aroudn the living room and you pick a fight with gravity, and gravity wins, and all of a sudden, an event that once upon a time would have left you sore for a few days instead leads to a snap, crackle, or pop. If that was your hip, it can quickly get very serious.

In a large series of older hip-fracture patients, about one in six died within the year after the fracture (Nature, 2019). Roughly 40% couldn’t walk without help when they left the hospital. A broken hip can be the first domino in losing your independence.

This letter is aimed to give you the knowledge and context to lower your risk of that outcome to the greatest extent possible.

The unfortunate reality is that breaking bones is common. I was surprised to learn that half of women over 50 will break a bone because of osteoporosis.

The screening system is built to catch it, but comes a little later than you might hope

So why doesn’t your doctor catch this early? Perhaps understandably, the system is built to find osteoporosis once it has already set in and once scanning results could change medical decisions (about potentially starting medicines). The official recommendation is a DEXA scan (the quick, painless X-ray that measures bone density) for every woman at 65, and earlier for younger postmenopausal women who have certain risk factors (USPSTF, 2025).

But hold that recommendation up against the timing we just talked about. The fastest loss happens in the five to seven years right after menopause, which for most women lands somewhere around their early 50s. By 65, that window is in the rearview mirror.

In other words, the scan that diagnoses osteoporosis is very often ordered years after the bone loss that caused it. That’s because, even if a scan at age 55 came back with evidence that bone loss was beginning (but not yet at a severe level that would justify medication), the answer for “what should you do” would likely be “lock in on stength training and nutrition” which is the same advice that you would be theoretically already operating under in the absence of the scan.

(For men, the picture is different: the USPSTF says there isn’t enough evidence to recommend routine screening, though some other guidelines suggest testing men over 70.)

You don’t have to wait for action on that schedule, which we’ll get into below.

The trouble with “just be careful”

My worry is that some of the most common “tips” you’ll hear bubbling around the internet can actually raise your risk over the long term. For decades, women with thinning bones were told to be gentle. Walk, maybe lift a light weight, don’t strain, and above all don’t lift anything heavy, the implication being that you’ll snap. And I suppose that isn’t totally crazy at face value, since falls and fractures are exactly what you’re trying to avoid.

But it fails to address the biological reality that your bones are living tissue, just like your muscles, and they follow the same set of principles: load it and it strengthens, leave it idle and it declines over time.

Remember thigmomorphogenesis in plants?

“Be gentle” ends up cementing a downward trend, when you may have actually had the potential to reverse course or, at the very least, to slow the rate of change.

In 2018, a trial called LIFTMOR tested this in exactly the women that “popular advice” has traditionally been afraid to let near a barbell: postmenopausal women who already had low bone mass or full osteoporosis. Twice a week, 30 minutes a session, for eight months, they did the movements that may seem intimidating at first glance. Heavy deadlifts, back squats, and overhead presses, worked up to more than 80% of their one-rep max, plus jumping and landing hard on purpose (Watson et al., JBMR 2018). They used real weight and trained with real intensity, and only for a total of one hour a week, split across two sessions.

Who do you think had better outcomes for their bone density? The heavy-lifting group, or the gentle-exercise comparison group?

Here’s what happened over those eight months. The lifting group gained about 3% bone density at the spine and roughly held steady at the hip, with no further loss.

The comparison group, doing gentle exercise, lost some bone at both sites. So the difference between the groups came out around 4% at the spine and 2% at the hip (Watson et al.). And that’s in just eight months! As you’d expect, the lifters’ muscles also got stronger.

Across those eight months of heavy lifting by those 43 women with fragile bones, there was one minor injury. The idea that these women were too fragile to train turned out to be untrue. Read the fine print, though. This was a small trial, every person was screened for red flags before starting, each was coached in person by a physiotherapist, and the load was built up gradually over weeks. “It was safe” means it was safe under those conditions. This is an argument for the safety and efficacy of coached heavy lifting. That’s quite different from improvising your own heavy program off a YouTube video without consulting your doctor.

But consider the difference between those two experimental groups for a moment. The “gentle exercise” group drifted further into osteoporosis. The heavy training group actually reversed some of the decline at the spine and stopped the loss at the hip, on just two half-hour sessions a week, doing the exact activity they may have been told would break them (with expert coaching).

Be real with yourself for a moment. Does your weekly routine put you closer to the “gentle exercise” group or the “heavy lifting” group? Think about what that means for your bone trajectory.

Your skeleton is responsive, today, to what you ask of it. It has no memory of what happened ten years ago. The cells in your bones and muscles respond to the circumstances and requirements you give them today. Load them up and they infer they’d better get stronger, in anticipation of future challenge. Let them rest and they won’t spend energy on growth they don’t seem to need.

Bone density, like muscle mass, is a number you can change. People have successfully improved it in closely monitored research studies. But most people never learn any of this until after they’ve already had a fracture, when the whole conversation has turned to recovery.

I’d much rather you learn it now, when prevention is still on the table.

Before you touch a barbell (read this!)

Two things of very high importance.

First, this is not medical advice for your specific body. In the rest of this letter I’ll describe a general “map” based on the published evidence. But not everyone holding the map should walk in the same direction, and it’s your doctor and your trainer who should help you decide what to actually do.

Second, where you are on the curve changes the plan. If you already have osteoporosis on a scan, or you’ve already broken a bone from a standing-height (or lower) fall, or you’ve lost height, or you have a known or suspected spinal fracture, you’re in a different category from the person trying to prevent all that. For someone in that circumstance, your doctor may want to discuss medication as a first move, and you should get your doctor’s clearance before you start doing heavy lifting on your own.

Heavy lifting can still be a big part of your story once you’ve been diagnosed and cleared by your care team. This letter leads with the physical training because most readers are earlier on the curve, in the osteopenia-and-prevention zone, where training does the most and your doctor may advise that drugs are nto yet warranted.

But whatever category you’re in: don’t start, stop, or change a prescription because of something you read online (including here). When you do start lifting, start light. A broomstick and a coach on your form come first, and loading up the barbell comes later. That kind of on-ramp is the whole point of doing this safely if you already have thinning bones.

What’s below the paywall

So far, we’ve discussed why you should care about bone health and why now is the time to act, namely that your bones are living, breathing tissues that respond to load.

The rest of this letter is the “how.” That content could be correctly and crudely summarized as “lift heavy weights and eat enough protein,” but actually carries about 3,500 more words of context below the line.

This part took me the most digging, because the nuance here is scattered across trial protocols and specialist guidelines and half-obscured under a mountain of supplement marketing. My paid subscribers are what enable my continued work on this letter while keeping a sponsor and an affiliate link out of the room.

Today:

• The lifting protocol, from a standing start: the key LIFTMOR movements, how to begin safely if you’ve never held a barbell, how heavy to go and how fast (or slow) to get there, and why “walking is enough” is a very expensive myth to believe.

• The food and supplement specifics:

  • the grams-of-protein target for bone (and the larger target if you also want more muscle gains)

  • why the “protein leaches calcium” scare is nonsense

  • how much calcium, food versus pills

  • the truth about vitamin D now that the guidance has changed

• A three-minute daily balance habit to head off the falls behind most broken hips.

• Your numbers, decoded: when to actually get a DEXA (and why the guideline waits until 65), what your T-score means, and how the free FRAX calculator turns it into your real ten-year fracture risk.

• A doctor-visit script, tailored to you: the exact questions to ask whether you’re a 35-year-old woman, a postmenopausal woman, or a 70-year-old man, plus the follow-ups to ask if medication comes up.

Subscribe now

A paid subscription also opens the full archive of deep dives, everything I’ve written and everything still to come. It runs $96 a year, which works out to $8 a month, or a little under $2 a week, and you can cancel in one click. My mission is to make it the most useful $2 you spend per week.

If even just one of the roughly 100 letters I’ll publish over the next 52 weeks gets you to start defending your bone health a decade earlier than you would have otherwise, or talks you out of an expensive supplement in favor of something that works, or gives you the words to raise this gently with someone you love, that’s the whole point of it.

Let’s start with a closer look at the LIFTMOR lifts that protected bone density:

Read more

-Morgan McSweeney, PhD (@dr.noc)

If you have browsed health headlines in the last two years, you have seen this headline in one form or another:

Popular sugar substitute linked to brain damage and stroke risk… ‘Natural’ Sweetener Erythritol Linked To Stroke Risk… This Ingredient Often Found in Protein Bars Was Just Linked to Potential Stroke Risk…

The ingredient in the crosshairs is erythritol, a “sugar alcohol.” It is the most common sweetener in keto ice cream, sugar-free syrups, and “zero sugar” protein bars, with its cousin xylitol close behind.

The stories are alarming and are built on published evidence. However, as usual, the headlines embody a small sliver of truth that has been run through institutional clickbait machinery to generate MaximumFear™.

Today we will take a few minutes to consider what the research actually shows. The current state of the evidence is neither “they’re obviously totally safe” nor “they’re obviously dangerous.”

Here is the big picture before we embark upon our quest for detail.

There is evidence for mechanisms of risk for sugar alcohols. But, it is nowhere near compelling proof that is pans out to actually increase risk in humans. The observational data are confounded by several interesting and important factors. As we will discuss in detail, the FDA has evaluated the research behind the scary claims, and has decided that there is not currently enough justification to act to reduce erythritol in our food supply.

In today’s expanded deep dive section:

• What did they find in the concerning studies?

• What was the FDA’s argument against taking regulatory action right now, after reviewing those concerning studies?

• (Opinion) How have these data (and the FDA’s argument) changed the way I think about sugar alcohols? Also, we must discuss the substitution effect.

• The study in healthy humans that increased my hesitation (for me, this differentiates sugar alcohols from the artificial sweeteners in diet sodas)

• What would it take to change my mind one way or the other?

This letter has no ads, no sponsors, and no affiliate links, and is fully reader-supported. In return, my paid subscribers get the extended version of most letters. If you prefer to conduct further reading on this topic yourself first: the FDA’s memo is public. I do not provide special access to secret material. What a paid subscription buys you is my analysis and my commentary as a filter.

Subscribe now

A paid subscription also opens the archive. A few recent reader favorites:

• Creatine: What Works and What’s Hype

• The Truth: Is Alzheimer’s Preventable?

• The Heart Attack That “Came Out of Nowhere”

It works out to $8 a month on the annual plan ($96/y), which is less than $1 per research letter. If it’s not useful, unsubscribe any time in one click.

Read more

-Morgan McSweeney, PhD (@dr.noc)

A machine recently measured the visceral fat wrapped around my internal organs and then, unprompted, congratulated me on it. The report said, in a little blue box: “You’re doing great!”

I have mixed feelings about being consoled by a body scanner, but these days I take encouragement where I can get it.

Here is what that supportive little software was so pleased about, and which is also the reason I’m writing this letter. Between two scans on the same machine, seven months apart, my weight went from 193 pounds to 188. Only about three of those were fat, and my muscle held about the same.

Now, if you were to look only at the total numbers there, you might say that nothing much happened.

And I would agree.

But what those total numbers don’t tell you is that my levels of visceral fat, the deep stuff packed around my liver and intestines and organs, the type of fat that actually predicts disease risk, fell by roughly 70%, from 0.52 pounds to 0.15 (from 15.3 cubic inches of visceral fat down to 4.3).

A reduction in visceral fat is indeed something to celebrate.

And part of the cool thing here is that you do not have to remake your body to move it considerably. In a rare, helpful quirk of biology, the harmful visceral fat comes off first, and it comes off in response to a short list of changes.

As a matter of principle, I must complicate my own n=1 example. DEXA estimates visceral fat with an algorithm that is shaky at low mass, so I trust a conclusion of “most of it came off, and fast” more than the exact “70% reduction”. I was already fairly lean, and I changed several things at once, so I can’t provide a single cause. But despite those caveats, I care a lot about the directional change in my visceral fat. I will show you why I think you should care about your visceral fat, too, and what to do about it.

Visceral fat is metabolically problematic in a way the fat under your skin is not

There are two broad kinds of belly fat. Subcutaneous fat is the soft layer just under the skin. Visceral fat sits deeper, wrapped around your organs. But the way it wraps around your organs is less of a “comfy hug” and more of a “metabolic python strangle.”

You see these two types of fat do NOT behave the same way. Subcutaneous fat is close to inert storage. Visceral fat is metabolically active tissue that leaks free fatty acids and pumps out chronic inflammatory signals, keeping the liver and whole body under an element of low-grade, chronic inflammatory strain. That is a part of why it tracks so tightly with insulin resistance, type 2 diabetes, and heart disease.

Is the systemic inflammation the fire, or just the smoke? Those inflammatory responses may be doing real damage on their own, or they may just be a readout of the deeper problem, with a direct interaction between the visceral fat and the organs driving the disease risk. I suspect the answer is that it’s some of both, with more emphasis on the latter. However, the exact answer doesn’t change the plan, because either way the fix is to get the visceral fat down, not to fret about the inflammatory markers themselves.

Subcutaneous fat (under your skin, which you can pinch), is much less related to your risk for chronic disease due to distinctions in metabolic profile, and that difference is exactly makes visceral fat easier to lose.

Visceral fat predicts diabetes, heart disease, and death EVEN AFTER you account for bodyweight

UK Biobank imaged fat depots by MRI in about 40,000 people and followed who got sick thereafter. Per one standard deviation more visceral fat, adjusted for BMI, the risk of new type 2 diabetes rose about 45% and coronary artery disease about 17%. The subcutaneous fat under the skin carried no added risk for either!

This distinction is central to this whole letter, so please pause to consider the implication of that pair of findings.

The visceral fat, specifically, is what I want you to focus your attention on.

Note that those percentage changes in risk are relative bumps on a low base, diabetes and heart disease each affected about 5% of this cohort, so a 45% bump from that ~5% starting point of absolute risk means a few extra cases per hundred. Still a meaningful difference in absolute risk.

Separate evidence has shown that waist size (a proxy for visceral fat) predicts death independently of BMI. In EPIC, a study of 359,000 Europeans, those with the largest waists had about double the mortality of the smallest after adjusting for BMI.

A pooled analysis of 650,000 adults found the same conclusion at every BMI, including normal weight: the men with the largest waists lost an average of about three years of life after 40, women about five.

These are observational data, and waist circumference is only a stand-in for visceral fat, but all the data I have found point in one direction for the health impact of visceral fat.

Down.

You can carry a dangerous amount of visceral fat at a completely normal weight

This is where visceral fat gets sneaky. In that same MRI data studies from the UK Biobank, a normal-weight man in the top group for visceral fat had a higher predicted absolute risk of type 2 diabetes risk (6.6% risk) than an overweight man in the lowest group of visceral fat (2.7% risk).

In other words, the assumption about future risk for type 2 diabetes that you would (unfairly) make based on looking at those two individuals with your eyeballs would be totally and woefully incorrect. This divergence between what you see and what is going on metabolically is critical to take a moment to reflect upon.

You can not tell how much visceral fat you have around your organs by looking in the mirror. This can be doubly deceiving for people who do not have overweight/obesity. Such a person can still harbor considerable visceral fat risk.

One “see a doctor” flag before we continue. If your belly is expanding fast, or out of proportion to how you’re eating and moving, especially alongside new purple stretch marks, easy bruising, or muscle weakness, see a doctor. Central fat can occasionally be the opening salvo from a thyroid, hormonal (think PCOS, recently renamed PMOS), or a medication problem that no amount of aerobic activity or berries will fix.

The fat on your hips and thighs is not driving your metabolic risk

In the same MRI study, fat on the hips, thighs, and backside did not track with diabetes risk.

So the target in the crosshairs today is visceral fat.

Not “fat.”

The good news: visceral fat is first on the chopping block

Now the reason for a little molecular celebration. Our biology doesn’t usually make things easy, but this is one case where it kind of does. When people lose a modest amount of weight, they typically lose visceral fat preferentially compared to the fat under their skin. A systematic review of imaging studies found this holds across diets, drugs, and surgery.

By “preferentially” I mean proportionally, not in raw pounds. There is simply more subcutaneous fat to go around, so you’ll probably lose more of that by sheer weight, but visceral fat wins on percentage. The meta-analysis that separates the two is clear on it.

…percent decrease of visceral fat was always greater than percent decrease of subcutaneous fat, with no differences between different strategies…

If you’re in the spirit for an anecdote (who isn’t?), I saw it in my own numbers, too. Whereas my total fat mass went down only about 5%, my visceral fat went down ~70%.

How did I get the visceral fat to go first? By sleeping with a slice of potato in my socks. (kidding)

It’s nothing I (or anyone) did - it’s because visceral fat is metabolically “hot”: it’s more sensitive to hormonal fat-releasing signals. Specifically, this is due to increased function of β3-adrenoceptors, decreased function of α2-adrenoceptors and increased ability of cyclic AMP to stimulate lipolysis in the visceral fat.

What that means is that more of the visceral fat comes off first, while the stubborn (and less risky) stuff on the hips and thighs holds on.

However, that difference in metabolic sensitivity is different from targeted “spot reduction” training. In other words, the visceral fat will preferentially come off at a greater rate than total body fat if you are at a modest total energy deficit or if there is some body recomposition going on, regardless of whether doing stomach-targeted exercises like crunches.

You don’t have to do crunches for this to happen.

A 5% loss shrinks visceral fat

In a graded weight-loss trial, a 5% loss cut intra-abdominal fat by about 7%, separately cut liver fat by 40%, and improved insulin sensitivity in liver, muscle, and fat, with participants in that particular study still classified as having obesity throughout.

For scale, five percent loss is 10 pounds on a 200-pound person. I lost about half that, didn’t lose any muscle or strength performance, and still moved my visceral number down measurably.

Increasing exercise reduces visceral fat even when total bodyweight does not change

You don’t strictly need the scale to move at all. A meta-analysis of 117 studies found that even in the absence of any weight loss, adding exercise was linked to a 6.1% drop in visceral fat, versus 1.1% for dietary changes alone.

The approach to reducing visceral fat is below.

The extended practical section is for the paid subscribers who make this deep research possible. Today:

• How to measure your own visceral fat for $5. The one tape-measure number that tracks your trend about as well as a $150 scan, how to take it right, and how often to check.

• The playbook. The diet change that beat a standard healthy diet three to one on the same calories, the sleep piece, and the training.

• Specific anti-visceral-fat nutrition options to start adding snacks, breakfast, lunch, dinner.

• DEXA vs. tape measure vs. the “visceral fat” bathroom scales. How much to trust the numbers, and what to ignore.

• The money pits to avoid. The belly-fat products and gadgets that do measurably nothing, with the studies.

• The drugs. Where the GLP-1s actually fit.

Subscribe now

Paid membership also provides access to all prior and future deep dive articles:

• Creatine: What Works and What’s Hype?

• The Truth: Is Alzheimer’s Preventable?

• The #1 Predictor of How Long You’ll Live

And here’s why the extended section is behind a paywall instead of an ad: I take no money from supplement companies, gym-gadget sellers, etc, and this letter is fully reader supported. That means I answer to your best interests alone.

On a subject where almost everyone handing out belly-fat advice is also selling you a hopeful tub of powder or a device, I’d rather you pay me eight dollars a month (that’s the average via the annual plan) for my unfiltered read of the data.

If it’s not useful, you unsubscribe in one click.

So: what moves it, in what order, how to measure it, and what to skip. Here’s the whole playbook.

Read more

Listen to our full discussion here on apple podcasts, spotify, or amazon music: https://lemonada.lnk.to/VO2MaxxingwithDrNocSG

In the mean time, I am now turning back to finishing up an article for tomorrow on visceral fat. (oh boy!)

-Morgan

-Morgan McSweeney, PhD (@dr.noc)

A while back I went looking through an old notebook for something unrelated and I found a note, in my own handwriting, that flatly contradicts something I now believe.

It was a small, triumphant note about how I had finally beaten the psoriasis that used to flare up very painfully on my hands. I had started taking oral Vitamin D right before we moved halfway across the country, the flare-ups faded not long after, and past-me had written it down with the confidence of a man who used his superior scientific wisdom to solve his own case: the Vitamin D fixed it. My weekly need for strong topical anti-inflammatory medicine from my dermatologist decreased from 5x/week to maybe once every other month.

But, as it would turn out, the Vitamin D did not fix it.

After two months of near complete symptom remission, in the spirit of science, I stopped taking the supplement, expecting my hands to revolt, and… nothing happened. They stayed perfectly fine and my need for my topical medication stayed at nearly zero, which means the supplement was very likely never the reason, and the credit belonged instead to some unknowable mix of everything else that had changed in the same stretch of time: a new latitude, more regular sleep, and the Norwegian 4x4 intervals I’d picked up that same month. Or, perhaps there was no external factor at all, and some biological process was going to shift regardless of what I did or did not do.

I had confidently pinned the whole thing on the one variable that happened to come in a bottle, and for which I had strung together a complex mechanistic argument for how it could work.

I stopped and chuckled when I saw that note for a second. Then I remembered it is completely normal, and went back to what I was doing.

(If you would like to subpoena me, my old notebooks are strategically scattered around our house. Check the big box in the attic next to the Christmas decorations. While you’re up there, please replace the lightbulb.)

Subscribe now

Ahead of a July 29 hearing, Senator Rand Paul, who chairs the Homeland Security committee, released more than 1,100 pages of Anthony Fauci’s private journals, covering December 2019 through December 2022. At the hearing, Fauci invoked his Fifth Amendment right more than a hundred times and declined to answer. Today, Republicans have moved to hold him in contempt, and that decision will now go before the entire Senate.

Within hours of last week’s hearing, both camps had declared victory. (What’s new?)

As a scientist, I want to talk about why I believe that everyone is making the same fundamental mistake about these notes.

First, what this letter is not. I am not telling you where I think SARS-CoV-2 came from. I am uncertain about the origin question, and I will not pretend a diary entry settles it any more than my own notebook entry settles the cause of my symptom remission. Nor will we discuss the failures or successes of public policy during the pandemic. This letter isn’t actually intended to be very much about Fauci or COVID at all. It is about the value of uncertainty.

None of this minimizes the pandemic. It killed more than a million Americans, it pushed hospitals past their breaking point, and I have little patience for anyone who waves the whole catastrophe away. The people who worked the front lines don’t need a senator’s document dump to tell them whether it was real. They were there, and many of them are still not okay.

The mistake is failing to realize that a scientist’s private notes and a scientist’s public statements are two different kinds of writing with very different purposes. A notebook is where the uncertainty is meant to thrive: the third hypothesis you can’t rule out, the result that doesn’t fit with everything else, the important finding that you can’t seem to replicate, and the thing you suspect at 2am but abandon by Tuesday. If these thoughts are not in a notebook, they are hopefully showing up in conversation with a colleague over coffee - - stepping outside the bounds of what is “clearly supported by the data” to posit on what might be true.

The public statement is where you have to commit to a best current estimate. But a scientist whose private notes matched their press briefings, word for word, would be a scientist who had stopped thinking somewhere along the way.

I run into this in my own work all the time. What an internal draft of one of my own grants or manuscripts says about a mechanism, hedged and arguing with itself, is not what lands in the polished final submission.

If you have never worked in a lab, you might be surprised how many routine assays come with their own unexplained quirks that everyone in the lab learns to respect. (”I don’t know why, but this one always behaves better on a cloudy day when it hasn’t rained…”) We write those down in the notebook precisely because we cannot yet defend them in public.

The private notes and the public documents rely on the same data doing two different jobs, and taking time to hold on to the uncertainty and challenge it and work through its implications in the private portion of its lifecycle is a critical part of the process of scientific thought.

If you forget that, it is easy to condemn hedging in private as proof of lying in public. I find that to be an error of category: private uncertainty is what public confidence is eventually built on. So, too, would it be an error to read the contents of private journals as vindication of being correct. Both would be making the mistake of treating a private sketchpad for reasoning as a source of truth about the world, when it is really nothing more than evidence about what one person was thinking on one particular day.

If you had recorded every discussion I’ve had in private with colleagues over the years, working through theories and analyses of scientific data, you could very well categorize me as being often wrong.

Or, perhaps nearly always wrong.

But, eventually, through good faith debate and critical examination of data, I update my beliefs over time.

We have ended up in the unfortunate circumstance where public accountability now runs through document dumps of private materials curated by someone’s political opponent, and yet we have agreed on no method for reading them.

As far as I can tell, nobody has measured a basic question, which is whether releasing someone’s private communications tends to productively change anyone’s minds or just harden the ones already made. We should probably know that before we do it again next month to someone else.

So, before someone dumps my (or anyone else’s) private notes for the world to judge, a few questions to ask before forming a hard opinion:

• Who chose the excerpts that would be released and highlighted, and what might they have left out?

• Was this a working document meant to process internal thoughts and uncertainty, or a communication meant for other people?

• What is the context in the pages that you never saw?

• Am I at risk of forming a strong opinion based on second hand reporting, or have I read the primary notes myself?

• Would I accept this standard if it were my notebook or my private conversations being judged, selectively dumped by someone who disliked me?

None of this is a claim that “both sides” here are equally right, or equally decent, or acting in equally good faith. I have my own opinions on that. I think that Paul & Co.’s belligerent pursuit of Fauci is unproductive and harmful for multiple reasons, but that’s beside the point of this letter.

Hopefully it is clear that I am making a narrower point about method and category: there is value to working through the uncertainty that is inherent to private notes and private conversations.

Acknowledging the different function of an uncertain private conversation or document vs. a best-current-guess public statement is what I would want applied to my own decision making, and what I would want applied equally to people I admire and people I cannot stand.

I have been wrong in private more times than I can count, and I will be again by Monday morning.

-Morgan

Morgan McSweeney, PhD

-Morgan McSweeney, PhD (@dr.noc)

For decades, the implication was that artificial sweeteners are inert: they provide the taste of sugar but then slip through you untouched and largely invisible to your body.

And, to be fair, that does hold up with respect to the health outcomes you probably are most concerned about: cancer, heart disease, preterm birth, early death, etc.

Consuming diet sodas or other products with sweeteners has not been shown (in high quality studies) to increase risk for any of those outcomes.

But if you take a moment to speak with your friendly gut microbiome, you will find that sweeteners are not totally inert. As you know, you are not the only one sitting down at the table to crack open a diet beverage or zero-sugar yogurt. You coexist with a friendly swarm of trillions of living cells, and you get to choose what they eat. (Oh the power!)

That landscape is where the artificial sweetener conversation becomes just a little bit murky.

Not in a “oh, this is concerning for my health” type of way, but in a “oh, I guess there is more uncertainty here than I realized.. I guess there’s a lot left to learn” type of way.

• What happens if you conduct a study where you feed mice some sweeteners, collect their poop, and then transplant their poop into other mice?

• What happens if you enroll some people in a study, feed them sweeteners, collect their poop, and then transplant their microbes into mice?

Well.. something happens. Today we will examine what’s going on in your gut when you consume sweeteners.

The full breakdown is below

I have no ads, no sponsor, and no affiliate products - this letter is fully reader-supported. Here’s everything in the rest of today’s letter:

• The results of the poop-bacteria-transfer experiments. (Yuck.. but cool?)

• A breakfast recipe that will have your microbiome giving you thunderous applause. (My foray into the world of writing recipes. Don’t get your hopes up too high! Simple to make, though..)

• An overview of the microbiome: what those trillions of bacteria actually do for you, and why “gut health” content online oversells its confidence.

• A few key studies that found nothing: key null results, and why the exact same sweetener has sometimes led to opposite headlines.

• Sweeteners as tiny drugs for bacteria: this is one of the more strange concepts, wherein sweeteners may help bacteria swap genes amongst eachother (… in vitro).

• The lab signals that researchers continue to debate: a sucralose impurity that damaged DNA in cultured cells, and a reminder about what that 100,000-person observational cancer study saw when it looked for an association with sucralose.

• The key lifestyle factor that overshadows the impact of sweeteners by a wide margin.

• (Commentary): On mechanisms vs. outcomes for the gut microbiota.

A paid subscription also opens the archive. A few recent reader favorites:

• The Truth: Is Alzheimer’s Preventable?

• The Heart Attack That “Came Out of Nowhere”

• Creatine: What Works and What’s Hype

Subscribe now

Here’s why I run my letter with a paywall rather than with advertisements: I take no money from any company and I don’t sell a gut cure, a probiotic, or earn a commission on anything. This is fully reader supported, which means I answer to you and you alone.

On a topic where most of the loudest voices are also selling you a probiotic or a fourteen-day “gut reset,” I’d rather provide my unfiltered read, free of commercial bias.

If it’s not useful, unsubscribe any time in one click.

Read more

Dr. Morgan McSweeney, PhD (@dr.noc)

“Artificial sweeteners speed up memory loss.”

That was the gist of recent headlines across The Hill, Harvard Health, and ScienceDaily, all taking aim at the same study.

Popular outlets just can’t seem to resist the temptation to pour gasoline on the tiniest spark of evidence about something that sounds frightening about brain health these days. I am sure they have the clickthrough data to back up that decision.

Many of the recent headlines quote the big finding: “62% increased risk of cognitive decline” associated with “high sweetener intake.”

They emphasize that kind of number because they know it will give you a little twinge of minor anxiety and make you look sideways at the drink on your desk.

It’s also the kind of number that should trigger your sniff-test: 62%? That’s a massive effect that should be easy to replicate in other studies. Why are we just hearing about this now?

So, with low expectations, I pulled the actual paper and read through it, tables and figures and all.

Perhaps now is a good time for me to say that I have no ties to the beverage industry, and I am not here to defend soda. I do not drink the stuff myself. Here I sit with a big cup of coffee and an even bigger glass of water. Today, I want to show you the difference between what they found in the paper vs. what you have seen in the headlines.

And just to be fair, they certainly did find a few things in the paper. That’s what makes this tricky to interpret at a glance, and valuable to walk through the biggest and most translatable examples. It’s not as if the entire paper was fraudulent or that it came back with null findings.

Here is what we will cover today:

• Where the “62 percent risk of cognitive decline” comes from, and the way that the headlines mislead you about the memory results.

• How to recognize the perils of this type of study design - whatdid it not account for?

• How much should worry, and what actually moves your brain health risk?

As you know, my letter runs no ads, has no sponsor, and pitches no affiliate links for you to purchase - this is an entirely reader-supported page, which means I answer only to you.

A paid subscription opens access to every deep-dive I’ve written, and everything still to come. If you’re ready to dig back, a few popular starting points are:

• Creatine: What Works and What’s Hype

• The Heart Attack That “Came Out of Nowhere”

• The Truth: Is Alzheimer’s Preventable?

• The #1 Predictor of How Long You’ll Live

Eight dollars a month (via the annual plan, which works out to ~$1 per deep dive letter), and you’ll walk away knowing exactly what’s worth your effort and what’s a marketing trick.

Subscribe now

If it’s not useful, unsubscribe any time in one click.

Read more

There is a clinical trial for a peptide on the U.S. government’s official medical registry, right now, marked “Recruiting,” that openly admits it is not real. The first line of its own description, and I kid you not, is this:

“This fictional study is an example of a ClinicalTrials.gov-style record.”

That registry, ClinicalTrials.gov, is supposed to be the honest ledger of medical research in this country. Doctors trust it. Reporters quote it. AI chatbots scrape it and hand it back to us as truth.

And this fake entry has been sitting there for months, long enough to get swept into the databases that sell trial data to the pharmaceutical industry, quoted back to me as real evidence by an AI chatbot, and written up by a wellness site as a promising development to watch.

Nobody, it seems, noticed the word “fictional” sitting at the top of the study description for that record nor the “(Mock Study)” disclaimer in the title of another.

I only found it because I was digging deep while looking for something else.

Here is the background.

This Wednesday I published a letter on the peptides that RFK Jr.’s panel voted last week to allow for compounding in the United States.

Three of the six peptides the panel backed were noted by FDA reviewers to have zero human data. So, I did the obvious next step and checked it out myself. I searched ClinicalTrials.gov to see whether anyone was even planning studies for these peptides in people.

For one of them, a peptide called TB-500, I found what looked like a real, ongoing trial: NCT07487363, a placebo-controlled study in cardiovascular disease, marked “Recruiting.”

Interesting!

Then I opened the full study description, and things got even more interesting…

The peptide listing that admits it is fictional

This is what you see when you first search “TB-500” on ClinicalTrials.gov:

It says it is recruiting, placebo-controlled, and being tested for atherosclerotic cardiovascular disease and endothelial dysfunction. That looks official at a glance, even to my trained eyes…because it is exactly the same info that would be provided by a real clinical study.

But what do you find if you click through to the study details?

“This fictional study is an example of a ClinicalTrials.gov-style record…. This example record models common ClinicalTrials.gov data elements….”

It would appear that an official ClinicalTrials.gov entry, marked as though it is Recruiting, is actually a fictional study of a peptide that, in the real world, has no published human trial data at all.

This is true mastery of irony.

Eight listings, one sponsor, one hospital, all marked as recruiting

Curious as to who was creating fictional records of clinical studies for peptides (and why), I searched further.

Turns out, that phony TB-500 clinical trial record is not alone. It is one of eight clinical trials, all of which were registered by the same sponsor (Hudson Biotech), all listing the same study site (Peking University Shenzhen Hospital), all first submitted between February and March of this year, and all currently marked “recruiting.” Between them, they claim 3,399 participants.

Three of the eight (in orange shading) note their own fraudulence: the TB-500 entry calls itself a “fictional study,” the Melanotan II entry opens “This example interventional study record describes…,” and the tesamorelin entry includes “(Mock Study)” in its title.

Two more (in red shading) are not new compounds at all, but strange carbon copies of Eli Lilly’s real GLP-1 peptide medicines, stealing the original trial names and even the company’s internal code names for the molecules.

Between the three self-labeled mock records and the two cloned Lilly trials, there is no question in my mind that the whole batch of 8 entries submitted by “Hudson Biotech” is bogus.

And that is exactly what makes the remaining three “normal” listings so misleading.

If you look at them on their own, the BPC-157, MOTS-c, and GHK-Cu entries read like ordinary trial records. Nothing in them says “fictional,” nothing says “example,” and they are not copying the code names from big pharma companies.

The only tell is it is part of the larger batch of 8 trials including 5 obviously fake ones from the same company.

The bogus clinical trial listings have already been picked up as “real” by data monitoring services

At least three commercial databases have taken in the data for these bogus trials, and are pitching the data through their “monitoring” services.

DataLookout presents it as a “Hudson Biotech Clinical Trial Pipeline, 8 active.”

Veeva’s trial portal also presents their status as “Enrolling” even for the trials with “fictional” in the description or “mock” in the name.

PatSnap Synapse, a drug-intelligence product sold to the pharmaceutical industry, has built a pipeline entry for “Pentadecapeptide BPC 157 (Hudson Biotech).”

All of the commercial databases treat these as a coherent pipeline from a real sponsor.

A product that resells public information on a company’s trials and never notices that some of those trials are fake does not strike me as particularly valuable.

I’ve come up with two explanations for this

I cannot say with certainty what would motivate “Hudson Biotech” to produce these listings. Two possibilities seem worth considering.

The first is that this is an attempt to game AI search and build an appearance of false authority.

I tested this with a minimal degree of rigor, asking an AI chat “is there a clinical trial for TB-500 (the one that calls itself fictional)? I was told there was:

“at least one registered clinical trial specifically involving TB-500… but that it was early-stage (focused primarily on safety and biological effects) and that no results have been published yet.”

and it gave me a link to that fictional trial listing!

If the intent was to spoof AI scraping bots, it’s working.

Could it be good for product sales?

Conjecturally, if Hudson Biotech is in the business of selling “research grade” peptides to enterprising “researchers” this type of AI-spoofing could be good for product sales.

After all, you’re going to be injecting yourself with unregulated peptides from overseas manufacturers, wouldn’t you want to use the peptides from the company that is running the actual “clinical trials?”

To check out this possibility, I went looking to see if I could purchase some peptides from Hudson Biotech. It soon became clear that I would have a hard time figuring out who “Hudson Biotech” really is.

They don’t seem to have a website.

I did find a company called “Hudson Research” that was very excited to sell me all of the peptides listed above for $50-$60 per vial “for laboratory use only.”

However, to be clear, I have no evidence that Hudson Research and “Hudson Biotech” are the same operation, or that whoever runs this peptide store has anything to do with the phony peptide study listings. “Hudson” is a common name, and the overlap could easily be pure coincidence.

Hudson Research Disclaimer: The information provided on this page is for educational and research purposes only. The products mentioned are sold as research chemicals and cosmetic raw materials, not as finished drugs or medical treatments. Not approved by the FDA for human consumption.

I managed to find a visual representation of the number of people who are convinced that these products are primarily being bought for the purpose of “research”:

Full of suspicion, I added a few vials to my cart and went to checkout to see if there was any safeguard in place to prevent individuals from purchasing their research products for non-research purposes. As you and I both expected, the answer was no.

In fact, in the shipping information section, they helpfully prompted me for my house and apartment number:

…and then presented me with diverse options for payment beyond a card, including Venmo (5% discount), cash on delivery (???), and cryptocurrency (10% discount).

It is well known in the scientific research industry that the finest research-grade chemicals are purchased with Monero and shipped to the researcher’s home address.

Regardless of whether “Hudson Research” (the peptide vendor) is related to “Hudson Biotech,” (the fictional peptide clinical trial sponsor) this is emblematic of the “gray market” surrounding the current availability of peptides with little-to-no published human data.

The second possible explanation is less sinister and… I think much less likely.

ClinicalTrials.gov has a separate test environment for practice submissions, at prstest.clinicaltrials.gov. Perhaps they meant to “practice” registering clinical trials, and accidentally used the live system instead?

I just can’t believe that. Eight records, submitted over three weeks, with two of them listing another company’s identifiers (and exact copies of their study details), is not a slip of the finger between the test environment and the public site.

For this piece, I tried to reach everyone with a stake. Via email, I contacted the National Library of Medicine, which runs ClinicalTrials.gov, and Eli Lilly, whose trial name (SURMOUNT-1) and compound code (LY3437943) appear on two of these records. I also reach out to the listed contact for study sponsor, “Hudson Biotech,” and the named study site, Peking University Shenzhen Hospital.

Only Eli Lilly responded to my request for comment. Per Lilly’s spokesperson:

“Lilly is aware of these registrations. We have no relationship with Hudson Biotech and did not authorize its use of our protocol identifiers or the SURMOUNT-1 name. We are committed to providing patients and physicians with accurate, trustworthy information about our clinical trials.”

As of the time of publish, I have not received any other responses. I will edit this post if I receive responses in the future.

What do you make of all this? I am curious to hear your thoughts, as I am not quite sure what to think myself.

Leave a comment

Either of the above explanations leaves a big problem

If this was deliberate, someone planted false evidence of pending human research on a federal registry, and it has worked well enough to fool AI chat models and to reach pharmaceutical intelligence platforms, convincing them to report that studies are Recruiting.

If the listings are due to sloppiness or some accident, then a batch that labels some of its own records “fictional” and “(Mock Study)” propagated into official and commercial databases and got written up by the peptide press… and nobody caught it for five months.

Either way, the official ClinicalTrials.gov registry does not appear to rigorously check validity of submitted materials, and the downstream products that resell registry data do not appear to, either.

As far as I can tell, no peer-reviewed study has ever estimated what fraction of the records in the clinical trial registry describe trials that never took place (or were never intended to be real trials in the first place).

Even if these were real trials, registering a trial does not provide evidence of safety or efficacy until the results are in.

If you want to read my longer piece on what last week’s FDA vote on peptides actually decided, and why “allowed to be compounded” is not the same thing as “FDA approved,” it is linked below my sign-off.

Best wishes,
Morgan

Archived links:

• fictional TB-500 listing: https://web.archive.org/web/20260730140902/https://clinicaltrials.gov/study/NCT07487363

• (Mock Study) tesamorelin listing: https://web.archive.org/web/20260730140936/https://clinicaltrials.gov/study/NCT07481734

-Morgan McSweeney, PhD

RFK Jr.’s panel wants to make minimally-tested peptide injections more widely available to Americans.

First, what are peptides? Small, biologically active molecules that are purported to have benefits such as enhancing muscle growth, promoting injury healing, or causing certain death (depending on who you ask) and they are often pitched on social media alongside convenient retail discount codes.

Indeed, individuals who have experimented with such peptides as BPC-157 or KPV are eager to share their personal experiences online.1

I have no doubt that many of these peptides have biological effects.

However, at risk of pointing out the obvious, “having biological effects” is not a compliment in the eyes of a researcher. There have been plenty of unexpected and undesirable outcomes as experimental interventions make the voyage from being “promising in animals” to “tested in humans.”

After all, if the peptide that makes your biceps Instagram-ready also happens to sprout you an extra big toe, you will have trouble putting on your Birkenstocks.

Some peptides are spectacular.

Anyone with the true spirit of science in their heart should remain open to the possibility that some of these peptides may prove to have favorable benefit/risk profiles after rigorous clinical investigation. Plenty of currently-approved medications are peptides, after all, including the GLP-1 drugs (e.g., semaglutide, tirzepatide).

But “after rigorous clinical investigation” is exactly the beans of the issue right now. The list of newer peptides considered last week simply have not been studied very well in humans.

Some of them haven’t been studied in humans at all!

The new peptides have very little (or no) human data

FDA’s scientific reviewers went looking for any published study in which TB-500, MOTS-c, or KPV had been given to a human being, and came back empty on all three.

However, the question in front of the FDA right now is not whether these peptides work, nor whether their benefits outweigh their risks for a specific indication.

Nobody is claiming to know whether these new peptides are safer or effective. That would be a question of FDA approval.

The present question is whether compounding pharmacies should be allowed to prepare them for patients with a prescription.

Those are two completely different decisions and the coverage keeps mashing them together. True - every one of these peptides was nominated to the FDA for a named medical condition: BPC-157 for ulcerative colitis, TB-500 for wound healing, epitalon for insomnia. But, confusingly enough, they are not actually voting on whether these peptides should be approved to manage those conditions, but whether to allow compounding of the molecules, which would just mean that a pharmacist in the US can legally weigh out the powder if someone has a prescription.

That is, as an alternative to what happens right now…

Currently, pioneering individuals with high tolerance for risk and low fear of biological uncertainty can purchase peptides from overseas manufacturers who will happily sell them to you “for research use only” (wink wink). Tales of the peptides’ properties float across the internet with stories of recovery, injury repair, sleep, and longevity.

If these peptides went onto the compounding list, that grey-market internet purchase would be replaced by something that looks a great deal more like actual medicine. A licensed prescriber writes a prescription for you by name, a licensed pharmacist weighs out the powder for you specifically, and you get a labeled vial. That’s pretty much it. The law requires a prescription and a named patient; it does not require a physical exam and it does not care whether rigorous clinical data have shown that the benefits outweigh the risks.

On whether to allow domestic compounding, I have come to see this as a question of harm reduction and of the value and role of regulation in industry. However, that must be weighed against the considerable shortcomings of the current compounding regulations.

Most drugs that start clinical trials fail

For 4.5 years back in graduate school, I worked on a problem almost nobody outside my field has heard of: how the immune system reacts to a common building block used in a lot of modern medicines. My job was to keep the body from treating an otherwise-useful drug like an intruder.

Me and Zina in my grad-school lab.

The approach we developed worked well in animals, extending the circulation time of the useful medications in mice and preventing the severe, anaphylaxis-type reactions that can (rarely) make an otherwise good therapy dangerous for some patients. I am proud of that work, and, in the years since, it has started to attract the support it needs to move closer to testing in humans.

But here is the thing. At the end of the day I have no idea whether it will work in people and whether its benefits will outweigh any potential risks. None. Even though I have spent years as close to that investigational intervention as anyone, I cannot give you a certain answer, and nobody can, until the human trials are run. Everything we saw in the animal data was promising, but a promise is not a result.

Only about 7.9% of drug development programs that enter Phase 1 testing go on to approval, across 9,704 programs tracked over a decade. Roughly 9 in 10 never make it: some fail on safety, some simply do not work very well, and some are dropped for commercial reasons.

So when somebody tells me that I should try a peptide because wellness clinics have used it for years and everyone says it helps, what I am hearing in my ears is a claim that in my field that ends up being wrong about nine times out of ten.

What happened last week

An FDA advisory committee spent two days last week voting on seven of the new peptides, and recommended six be considered allowed for compounding with a prescription. Only emideltide failed, and it failed by a single vote, 6 in favor to 7 against.

Note that the appointed advisory committee making these votes did not consist of the FDA’s own scientists. The career FDA scientists who wrote the review documents to prepare the scientific briefing for the meeting recommended against all of them, in consecutive sentences of the form:

“FDA is proposing that [substance] NOT be included on the 503A Bulks List.”

As you have probably assumed by now, I would not personally choose to inject myself with these peptides in the absence of compelling data on safety and efficacy in humans.

But who should get to decide that?

Should that be the individual’s choice to take on that risk (or not)?

This letter continues below.

My paid subscribers are the reason I can dedicate time to writing this letter: as you know, this letter has no ads, no sponsors, and no affiliate links. In return, the extended section of each letter is just for paid subscribers. Today, we discuss:

• What did they find when people have actually tested the peptides people buy online. Surprise: the quality of “research grade” peptides sold online by overseas manufacturers is about what you would expect.

• RFK Jr.’s argument for letting pharmacies do this, vs. the practical considerations that have kept me from getting on board with that argument.

• Shouldn’t people just get to choose for themselves? Why does the FDA exist at all? A troubling account of what can happen in the absence of regulation (leading to the death of dozens of children in 1937).

• What a peptide program looks like when it is done properly: what semaglutide, tirzepatide etc. have to prove.

Subscribe now

A paid subscription also opens the archive. A few popular deep dive places to start reading:

• The Truth: Is Alzheimer’s Preventable?

• Creatine: What Works and What’s Hype

• The Heart Attack That “Came Out of Nowhere.”

$8 a month on the annual plan (which works out to about $1 per deep dive letter).

If it is not useful, unsubscribe any time.

Read more

The email from “S.”

“Where is your background, expertise, resume etc? Therefore, I can make an informed decision about subscribing.

Thank you for your response in this matter.”

Quite a good question.

Actually, I’m a little embarrassed to have failed to provide any introduction here at all, aside from an occasional anecdote as part of a letter.

If someone is going to provide education on important topics in health/science, you ought to be able to check whether they are who they say they are.

After all, if there was a requirement for warning signs before browsing the “Health & Wellness,” “Health Politics,” and even some of the “Science” charts of Substack, you would be cautioned to Slow Down to max of 20 mph as you enter… the “Quack Zone.”

Here are the “verification” links I sent back to S.

I invite you to make your own assessment of my likelihood of being built of flesh and bones (rated out of 10, providing rationale optional).

Leave a comment

• (3rd party vetting) A New York Times profile on my work, with such varied accolades as:

  • “..he’s a real human being..”

  • “Dr. McSweeney became mildly famous for debunking misinformation…”

  • “He doesn’t act like a nerd on camera, he just is one — no bow tie necessary.”

  • “With a style somewhere between a high school chemistry teacher and a youth minister, he has made a name for himself with silly dances and vaccine explainers. ‘He’s sort of like a new Bill Nye’…”

• My academic publication list on PubMed: (link)

• My LinkedIn: (link)

• Talking to someone you know is real: And if you’d like to hear me talk at length, unscripted and without notes and in person (which I think is probably the easiest way to confirm I’m not an AI-generated character), here’s a long video podcast I recorded last year with Neil deGrasse Tyson & co.:

I have a PhD in pharmaceutical sciences / immunology from UNC Chapel Hill’s Department of Pharmacoengineering and Molecular Pharmaceutics, I work in drug development (in biotech). My specific areas of interest within the fields of pharmaceutics are more on the side of “biology” and less on the side of “chemistry,” with projects related to the immune system, inflammation, animal models of toxicity and efficacy, and antibody-mediated medicines for applications ranging from respiratory infectious disease to non-hormonal contraception.

We have one daugther and a dog (who occasionally feature in letters), and one of my favorite hobbies is going for a walk or jog on a forest trail.

I started doing this work because I got tired of headlines and influencers scaring folks about medical topics with misleading hot takes, all to drive them to click a link or buy some affiliate supplement product.

I would like to learn more about you at some point, so I must figure out how to set up a reader survey. I’ll attach that idea to a dart and throw it at my to-do list.

Until next time,
Morgan

Morgan McSweeney, PhD (@dr.noc)

Dr. Morgan McSweeney, PhD (@dr.noc)

Today we are talking about the new blood tests that claim to forecast your odds of cognitive decline, including whether they can really see your future, what those frightening percentages actually mean once you look closely at the study, and what you are supposed to do with an answer that no one can “treat” yet. I have seen many articles from popular outlets on this, and they all have been at a “surface” level of coverage.

I am writing this with the depth and nuance I would want communicated to my own parents (who will, in fact, read this letter. Hi!)

This will not be a short read, but my responsibility to you is that it should be the most useful thing you read about your brain health this week.

Most weeks, I take our daughter to the science museum near us.

There is a butterfly exhibit during the warmer months - a warm, double-doored room they keep stocked with butterflies that emerge from chrysalises shipped up (via Fedex) from Central America. They give you a laminated identification sheet on the way in, and my daughter takes it seriously, working down the little pictures to name each one she spots drifting past.

There is one big rule: no touching. She follows that rule to the letter. Nobody, however, said anything about talking. So, when we come to a butterfly that has landed on a fruit feeder and refuses to move, she leans in close and tries to motivate it into the air, first with a polite “FLY, BUTTERFLY!”

And then, when that inevitably fails to motivate the butterfly to action, by pretending to be a tiger.

Her favorite butterfly in the exhibit is the blue morpho. Its wings are a plain, leafy brown with some “owl eye” circles when they are folded up at rest, and then it opens them and there is this shock of electric blue on top that was hidden the entire time.

Blue is her favorite color, so the morpho is her favorite butterfly, and when we find one sitting still with its wings closed to conceal all that hidden blue, she does everything a two-year-old can think of to coax it into showing what she knows is hiding inside.

The biggest news from the last month in Alzheimer’s research is a set of blood tests that the headlines promise will show you something hidden inside your own head, years before it would ever surface on its own.

The important question underneath all the buzz, however, has not been thoroughly addressed:

Who is this relevant for, and, once you have coaxed the hidden thing into the open, then what?

A blood test now claims to see “Alzheimer’s” coming a decade before symptoms

At this year’s big Alzheimer’s meeting in London, and in a paper published in JAMA, researchers reported that a single blood draw can estimate your odds of sliding into cognitive impairment over the next two, five, and ten years. The test measures a form of the protein “tau,” called p-tau217, that rises in the blood as the amyloid and tau pathology of Alzheimer’s builds in the brain, often long before a person notices anything wrong.

Here, there are two claims that are being bundled together that are actually quite important to separate. One, is that these tests can help diagnose people who already have memory problems. The other is that they can predict the future for people who feel completely fine and have no symptoms. The first is on more solid ground. The second is far shakier, and even a correct answer would leave you with the question of “what do I do with this information?” We should keep those two scenarios (people with cognitive symptoms vs. no symptoms) separate, because I’ve been reading a lot of confusing takes that merge the two together.

There is one more piece of precision that I have been surprised to see most popular coverage drop. The study did not actually measure who went on to develop Alzheimer’s dementia. It measured who crossed into cognitive decline of any kind, a category that includes milder trouble and causes that have nothing to do with Alzheimer’s.

The blood marker is specific to Alzheimer’s pathology. But the outcome it was scored against as a predictor is much broader. Some of that cognitive “impairment” is mild cognitive impairment, which is not dementia, sometimes stays stable for years, and, sometimes, even reverts back to normal.

So “78 percent risk of cognitive impairment” is not quite exactly “78 percent chance of Alzheimer’s dementia.”

If you already have symptoms, this test helps your doctor get closer to the answer

The recent test advances are good news. When someone shows up with symptoms of memory trouble, telling apart Alzheimer’s from the dozen other things that look like it has always been hard, and the confirming tests, like a PET scan or a spinal tap, are expensive, uncomfortable, and unavailable to many people.

At the London meeting, one group reported that when primary-care doctors were given a patient’s p-tau217 result (from the new blood test), their diagnostic accuracy jumped from 65 percent to 93 percent (specialists went from 74 to 89), and clinical management changed in about half of cases.

Two such tests are already cleared in the United States, and here is the fine print you need: they are cleared only to help evaluate people who already have symptoms.

These tests are not cleared for screening healthy people with no symptoms, and the mail-order “direct-to-consumer” versions a worried person can buy are an entirely different beast.

A word of caution on these tests. A positive result tells you Alzheimer’s pathology is present. It does not prove that pathology is what is causing the symptoms of memory trouble, and in older brains more than one problem is usually going on at once. For context on how common this pathology is: by the time people reach their seventies, roughly a quarter (23.5% per one estimate) of those with no symptoms at all already carry Alzheimer’s changes in their brains, as indicated by p-tau in their blood. Plenty of those people would have died at a ripe old age never knowing about the brain changes and never having symptoms of cognitive decline.

A careful workup still has to rule out the things that can mimic dementia but are sometimes treatable or even reversible: depression, thyroid problems, low B12, sleep apnea, alcohol, and a long list of common medications. Truly reversible dementia turns out to be uncommon, but a good workup still screens for it, because a blood result that pulls a doctor’s attention away from a treatable cause could do real harm. So my word of caution is this: the test is a real step toward a faster, better-informed answer, as long as it is used in collaboration with your physician and stays one piece of the picture, rather than being treated as the whole verdict and explainer for observed symptoms.

From their press release: To guide clinicians in the appropriate use of blood biomarker tests, the Alzheimer’s Association is preparing a concise decision-making tool to be posted on ALZPro®, its central hub of professional resources. The Association has also published Clinical Practice Guidelines that provide clear, evidence-based, brand-agnostic recommendations to support appropriate use of blood tests in specialty care in people experiencing dementia symptoms.

Selection of tests and interpretation of test results are tricky decisions to make alone, and they should be coordinated in collaboration with your doctor.

The bigger and more sneaky leap: applying that same test to healthy people with no symptoms

Now, this is where things become complicated. Taking a test that was initially built to confirm disease in someone who already has symptoms and pointing it instead at someone with no symptoms at all, to forecast whether they will develop disease years from now, is a different thing with a much higher bar to clear.

That is what the new JAMA paper is really about, and it is full of both promise and peril for the currently-healthy individual.

Should you get tested right now?

The “right” answer is determined in discussion with your doctor and may change entirely depending on which of these people you are.

If you have cognitive symptoms and a doctor is already working you up, this test is a new tool that may help solve your puzzle. Ask whether your clinician is using one of the FDA cleared clinical tests and how they will confirm a positive result if you get one. Ask about their plan to rule out other potential causes of memory symptoms.

If you feel fine and have no cognitive symptoms, and are asking because of personal concern or perhaps because a parent had dementia, the position of the researchers building these tests is a firm: “don’t take this test yet.” Current guidelines recommend against testing people without symptoms outside of a research study for a reason. In a moment I will show you the numbers that explain why. In brief, the positive predictive value among the healthy general population will be lower than among the people they tested it on in this paper.

If you already bought a direct-to-consumer test and are now staring at a result, there are a few important notes before you read too much into it. The first is that there is no single, standardized “p-tau217 test.” Different companies use different methods and different cutoffs, so, confusingly enough, a “positive” from one test could have been “negative” on another, and the answer you are seeing may not line up with the research numbers everyone is quoting. A mail-order result does not come with anyone to help you decide whether getting a test is appropriate in the first place, or to help interpret the answer afterward. Critically, the “accuracy” of this type of test when it is run in people with an average prevalence of disease will not be the same as it was in the published study, which was an enriched cohort. I will show you how that calculation shakes out in the detailed section below.

By the way, I just took a look online and found a few options for at-home blood tests for p-tau, ranging from $388 (with insurance) to $500-$900 if paying cash. There is no doubt in my mind that these companies are preying upon people’s fear of dementia to sell them these blood test kits, despite unanimous recommendations against this testing being done for healthy people with no symptoms right now.

This is a complicated enough question that no one should be expecting to run their own diagnostic workup at the kitchen table. This is something you should discuss with your doctor.

One more word before we move on, because I suspect some of you are frightened about your own memory right now, maybe because someone you love was just diagnosed and now each of your own misplaced words suddenly feel like a whisp of smoke. If that is you: grief, stress, and a stretch of poor sleep can produce exactly that kind of fog, the missing word, the walk into a room with no idea why you came. In someone frightened and exhausted who is now acutely aware of dementia, that may well be the cause. Raise it with your doctor, but in the mean time be gentle with yourself, and know that the fear itself is not proof of anything.

A note: I am a scientist, not your physician, and none of this is medical advice. The point of this letter is to educate you on the context that is missing from most coverage on this issue and to help inform the questions you bring to your doctor, so you can make the most of the fifteen minutes you get with them.

A few reasons to NOT be more concerned than you were a month ago.

Before we go deeper, let’s cover some of the less alarming facts the headlines tend to skip (I suspect precisely because fear gets more clicks and ad revenue than reassurance), as well as one consideration that may be time-sensitive.

First, the near-term rates of progression to cognitive impairment in the study were low across the board. In the study, the two-year odds of developing cognitive impairment were between 1 and 4 percent for everyone, whether their p-tau217 was low or sky-high. This is not a test that says something is about to happen to you very soon.

Second, a positive result can simply be wrong. The test is not perfect, and several ordinary factors can increase a healthy person’s blood levels of p-tau217 all on their own, further contributing to false alarms. The biggest culprit is reduced kidney function, which affects about a third of people over 65 and raises the number whether or not any Alzheimer’s pathology is present. Older age tends to push it up too, in large part because kidney function declines with age. The number can also be raised by obesity, a recent illness, and even how the blood was drawn and shipped. A single positive result may prompt your physician to confirm with a more definitive scan or spinal fluid, but it does not provide a diagnosis on its own.

Third, and this one may be time-sensitive: you may want to think about certain kinds of insurance before you test, not after. I am not an expert here, so take this as a lay word of caution rather than legal advice. Because a p-tau217 result is a protein in your blood and not a genetic test, the federal privacy law GINA, may not cover it. My understanding is that the Affordable Care Act would still stop a health insurer from using any such finding against you as a pre-existing condition with respect to health insurance. However, I am not sure that protection extends to long-term-care, life, or disability insurance. If those types of insurance policies matter to you, you should talk to someone who actually knows this area about whether you should get those policies before pursuing testing.

The things you can actually change today

So what can you actually do about your risk? Quite a lot, as it happens, and none of it requires getting tested at all.

This is the opposite of a scary result you cannot act on, and it is a good idea whether or not you ever get tested.

Like we spoke about last Saturday in the deep dive “The Truth: Is Alzheimer’s Preventable?”, the 2024 Lancet Commission on dementia estimated that up to 45% of dementia cases worldwide are potentially preventable or delayable by addressing 14 risk factors across a lifetime:

1. hearing loss

2. high LDL cholesterol

3. less education early in life

4. depression

5. head injury

6. physical inactivity

7. diabetes

8. smoking

9. high blood pressure

10. obesity

11. heavy drinking

12. social isolation

13. air pollution

14. untreated vision loss

Your risk is not a fixed number. Addressing those factors is precisely how you push it down, which turns “there is nothing you can do” into… well… a to-do list.

A word of acknowledgement: several of these factors, from education to air quality to hearing and vision care, are shaped by money and circumstance as much as personal willpower, so this is partly a public problem and partly a personal one.

The short version is that the same boring fundamentals that protect your heart protect your brain, and that is where a positive test, or just a worried mind, should send you - to double down on taking care of the risk factors that we already know matter for cognitive health, and which you actually have some degree of control over (unlike your genetics).

You do not need a diagnosis to start.

The questions to bring to your next appointment

If this is on your mind, here is a suggested list of what to ask your doctor, rather than ordering a p-tau test off the internet.

• Should I get tested at all given my current condition or would you watch and wait?

• If a result came back positive, what would change in my care this year or in the next few years?

• Would you confirm a positive blood test with a scan or spinal fluid? How would you rule out the less common reversible causes of cognitive decline?

• (And here’s an important one for everyone): of the 14 modifiable risk factors for dementia from the list above, which do I actually have? What are my blood pressure, LDL, blood sugar, and hearing and vision status, etc., and how are they trending, and which should we work on first?

That last question is the one that can actually change your odds. The blood test, as it stands today, mostly just changes how much you know.

Where to find support: if this topic is sitting particularly heavy on you or someone you love, you do not have to carry it alone, and you should not. The Alzheimer’s Association runs a free helpline, 24 hours a day, at 800-272-3900, with real people who can talk through testing, care, and support.

Everything above is free, and it will stay free.

Below the line is for paid subscribers.

I have no ads, no sponsor, and nothing to sell you here other than my time spent interpreting the published evidence. On a topic where nearly everyone writing is also selling you a “brain-support” supplement bottle or a test kit, your subscription is what allows me to provide unconflicted analysis. Here’s everything in today’s letter, with every study linked so you can check my work:

• What the frightening numbers actually hide. The “reassuring” 40 percent floor is inflated by the same factors as the scary 78%, because the group in the study is not exactly a random slice of the population.

• A clear demonstration of the challenge with interpreting a “positive” test result outside of the study. Why a test that is “90 percent accurate” in the study is not necessarily the same for the general public, with a direct example of how it can become hard to interpret a “positive” result when disease prevalence in the tested group goes down.

• The gene result you might have not asked for. Some of these tests report your APOE status, the Alzheimer’s-risk gene. The privacy law you are counting on may only protect a gene result only for health insurance and your job, leaving life, long-term-care, and disability insurance as potential risk considerations. A few things to think about before anyone takes a test.

• What a high test result really can tell you, and the data that I am excited about coming soon. Further exploration of the “then what?” … why there is still no treatment for a healthy person with a high test score … why a quarter of cognitively healthy 70-somethings already carry this pathology … and the important prevention trials finishing within the next few years.

By the end, my goal is for you to be able to look at the next frightening headline, or sit across from your own doctor or your own worried parent, and feel confident that you understand what to do or what questions you should ask.

A paid subscription also opens the full archive: every deep-dive I’ve written, and everything still to come. If you’re ready to dig back, a few popular prior articles are:

• The Heart Attack That “Came Out of Nowhere”

• The #1 Predictor of How Long You’ll Live

• Creatine: What Works and What’s Hype

Eight dollars a month (via the annual plan, which works out to ~$1 per deep dive letter), and you’ll walk away knowing exactly what’s worth your attention and what’s a marketing trick.

If it’s not useful, unsubscribe any time in one click.

Subscribe now

Let’s go one level further. The headlines are all about the risk of cognitive decline for the high p-tau group.

But what can we learn from the risk of those people with low levels of p-tau?

Read more

The scare started with an outright fraud involving 12 children. Researchers across the world then checked in more than a million more children and found nothing. But if you once watched your own child change right after a shot, that fear makes so much sense.

Morgan McSweeney, PhD (@dr.noc)

I’m an immunologist, and I still get this question every week, usually from a parent, often a little embarrassed to be asking it: could there be any truth to the link between vaccines and autism?

In late 2025, RFK Jr.’s “CDC” rewrote part of its own website to say that “vaccines do not cause autism” is “not an evidence-based claim,” adding institutional confusion to an already-confusing topic. So here is what we actually know, from the original sources.

It started with one paper, and that paper was a fraud

All of this traces back to a single fraudulent 1998 paper in The Lancet. Andrew Wakefield, a gastroenterologist, reported on twelve (12) children with a supposed new condition: bowel inflammation paired with developmental regression. The paper’s real claim was a gut disease (Wakefield later named it “autistic enterocolitis”) with the MMR vaccine floated as the hypothetical trigger. Twelve children, no control group, and a causal story built on parents’ recollections of when symptoms began. That was the spark for what has developed into a worldwide health concern.

Here is what came out afterward, mostly through the dogged reporting of journalist Brian Deer. Wakefield had been paid £435,643, plus expenses (eight times his reported annual salary), by lawyers who were trying to build a lawsuit against vaccine manufacturers, and those payments started two years before the paper was even published. None of it was disclosed. When Deer laid the paper against the children’s actual medical records, his verdict was blunt: “No case was free of misreporting or alteration.” And The Lancet fully retracted the paper in 2010. Britain’s medical board struck Wakefield off the register that same year. A BMJ editorial by its editor-in-chief Fiona Godlee and colleagues called the paper “an elaborate fraud.”

So the foundation of the entire vaccine-autism idea was a single paper that was found to be fabricated, by a man who was being paid huge sums of money to reach that conclusion.

Then, researchers checked, in more than a million children

After Wakefield, scientists around the world did the thing that he never did: they looked at large groups of real children and compared the vaccinated with the unvaccinated in big datasets with rigorous statistical methods.

In Denmark, one study followed 537,303 children and found the autism rate in vaccinated kids was essentially identical to the rate in unvaccinated kids.

A later Danish study of 657,461 children reached the same result. A 2014 analysis pooled data on 1,256,407 children across ten studies and again found no link between MMR and autism.

“The study strongly supports that MMR vaccination does not increase the risk for autism, does not trigger autism in susceptible children, and is not associated with clustering of autism cases after vaccination. It adds to previous studies through significant additional statistical power and by addressing hypotheses of susceptible subgroups and clustering of cases.”

These are some of the largest investigations ever run on any medical question, involving millions of children.

One important study in that analysis looked specifically at children who already had an older sibling with autism, (i.e., the kids at the highest genetic risk). If vaccines were pushing susceptible children over some “edge,” that is exactly where you would expect to see their effect. They did not. Even in that high-risk group, vaccinated and unvaccinated children developed autism at the same rate.

“But I watched my child change after the shot”

When you have watched something happen with your own eyes, large studies conducted in millions of other children can feel beside the point. For many parents, their child was hitting milestones, and then, perhaps sometime in the second year, the progress stopped, or words started slipping away, exactly around the time of a round of shots. When you have lived that experience, a chart of half a million children doesn’t seem to matter very much.

A sizable share of children with autism spectrum disorder, around 30%, go through what’s called regression, where they lose skills they had already gained. The average age when parents first notice potential symptoms of autism is between 1 and 2 years old, with wide variability across families, and formal diagnosis comes a little later. If you consider the recommended vaccine schedule, that age window overlaps with the rough timeline for several vaccines. For example, the first MMR dose is typically given at 12-15 months.

Although those two clocks overlap almost perfectly, they overlap for completely unrelated reasons. One is when a particular kind of autism tends to become visible. The other is simply when we protect kids against measles and other pathogens. The inevitable problem is that if you follow millions of families through that same time window, by chance alone, some children will regress in the days or weeks right after a shot.

To any individual family, that overlap feels exactly like gutwrenching cause-and-effect even if they intellectually believe that the “data” say the shot had nothing to do with it.

Vaccines and Miscarriage

The same disconnect between a vivid personal experience and “the data” exists for any family who has experienced miscarriage sometime after a vaccine. Unfortunately, miscarriage is common. Somewhere around 10 to 15 percent of recognized pregnancies end in miscarriage (with large differences depending on maternal age). When you count the very early losses detectable only by hormone testing, total post-implantation loss is closer to 31 percent.

That baseline risk is the number any vaccine effect has to be measured against, because in a population of millions of vaccinated pregnant women, a very large number of miscarriages will occur no matter what, purely by chance. The job of a rigorous safety study is to figure out whether the vaccinated group has more than that level baseline risk.

Everyone agrees that makes sense on paper, but that “logical” calculation against baseline risk can feel entirely invisible to any individual family who has experienced a loss.

Deconstructing the links that people send me.

Every day, I am forwarded messages and links from concerned parents and grandparents, asking “is this true”? Sometimes, it is a list of 20+ “peer reviewed sources,” that is floating around, assembled to appear like an incontrovertible evidence base demonstrating the risks of vaccines to children.

Though I do not have time now to go through dozens of those papers in detail (nor do I think you would find it particularly interesting to read through nerdy methodological criticisms of dozens of low quality papers), there are a few themes that I routinely find when clicking through the links that people send me.

The place to start is the phrase: peer-reviewed.

Most people hear that as a guarantee, a panel of experts who checked the work and signed off. What it usually means is considerably less impressive than you might hope. An editor sent the paper to two or three other researchers, reading in their spare time and almost always unpaid, who judged whether it looked competent enough to print. If you’re lucky, you get one or two reviewers who really try to do a good job. They are weighing whether the work seems plausible, and are certainly not re-running the experiment or even auditing the raw data for fraud. Wakefield’s paper was peer-reviewed. It ran in The Lancet, one of the most respected journals in the world.

That label of “peer review” covers a far wider range of quality than most readers imagine. There is an entire tier of journals that will publish almost anything in exchange for a fee, with review that runs from cursory to essentially none. A few years ago a journalist wrote up a cancer study rigged with obvious blunders (the kind any competent reviewer should catch within minutes) and submitted it to more than 300 of these pay-to-publish journals. Over half accepted it. A different journal accepted a paper whose entire text was a single sentence, “Get me off your f****ng mailing list” repeated many times, sometimes as illustrations or diagrams (see below).

The unfortunate realist is anyone with a credit card and an agenda can conjure a “peer-reviewed study” more or less on demand, especially in today’s era of AI-generated slop.

That is exactly why a lone paper, standing on its own, is such a weak thing to build a decision on. What actually persuades scientists is replication: the same answer surfacing again and again, in different countries and different groups of children and with different study methodologies, produced by teams that would have been delighted to find the opposite. That is what sits behind the autism research that shows no connection to vaccines.

The problems in the papers I’m sent tend to come in a handful of recurring tropes.

The first is leaning on self-report databases (like VAERS) that were never built to prove anything.

Governments run open safety systems where anyone can file a report that something happened after a vaccine. A physician can file one, and so can any parent or passerby who saw a post online. Nobody verifies these reports before they go in, and each is only a record that two things happened in sequence, not that one caused the other. To make exactly this point, a doctor once filed a report that a vaccine had turned him into the Incredible Hulk. It was accepted and logged like everything else, and only came out when the staff phoned to ask about it. You cannot even calculate an actual risk from these tallies, because no one knows how many uneventful vaccinations passed by unmentioned for every alarming report that got typed in. Yet a favorite move is to add up the frightening reports and present that raw count as though each were a confirmed, vaccine-caused injury. Those safety-reporting databases are specifically intended that way – to function as a broad net to catch any potential uptick in problems – but those signals are only the starting point for further, more rigorously controlled statistical evaluation and follow-on study. The problem arises when people point to the self-report databases as proof of causality.

The second is a missing comparison group. If you want to know whether a vaccine causes some harm, the only question that counts is whether vaccinated children experience more of it than similar unvaccinated children do. A paper that merely catalogs bad things that happened to vaccinated kids, with no one to measure them against, cannot answer that, however large the pile of anecdotes grows.

The third is mistaking timing for cause. In a country that vaccinates millions of children on a schedule, some of them will, by chance alone, fall ill or lose skills in the days after a shot. If you line those cases up the pattern can look compelling. It is the same coincidence we walked through with regression and with miscarriage.

The fourth is “fishing.” Compare enough groups and slice the data enough ways, and sooner or later one slice will cross into “statistically significant” on luck alone. If you publish that slice but leave the rest unmentioned, you can spin a headline out of noise. A notorious vaccine-autism “reanalysis” of the past decade did precisely this, coaxing an effect out of a single narrow subgroup without appropriate statistical control (and was later retracted).

These types of methodological flaws take training to recognize and interpret. My point is that a study’s mere existence, even a peer-reviewed one, carries far less weight than you would hope.

Most of these papers do no harm while they sit in obscure journals gathering dust. The trouble starts the moment a title and a scary “conclusion” get screenshotted and forwarded into a group chat with every caveat stripped away, where they travel faster and feel truer than any boring methodological critique I can write.

When the worry moved to mercury, we got a natural experiment that proved it was not an issue

When the MMR link kept failing to show up in large population studies, the concern shifted to an ethyl mercury-based preservative called thimerosal, which was used in some other childhood vaccines. Regulators acted on the precautionary principle: by 2001, thimerosal had been removed from or reduced to trace amounts in the routine early-childhood vaccine series.

Interestingly enough, that removal gave us something close to a real-world experiment. If that preservative had truly been driving autism, the rates should have flattened or dropped once it was removed from vaccines. Instead, they kept climbing, year after year, straight through the 2000s and up to today’s figure of about 1 in 31 children. Much of that rise (perhaps ~60%) traces to something far less sinister: we widened the definition of autism and got much better at spotting it. It remains unclear what specific factors are responsible for the apparent increase in autism diagnoses beyond changing definitions and detection. However, studies in millions of children clearly show that vaccines are not part of that answer.

Why this one isn’t a harmless belief

Continued fear around vaccines and autism carries a cost, and we are watching it come due.

Measles is one of the most contagious diseases we know of, and the MMR shot is one of the best tools we have ever built against it, roughly 97% effective after two doses. (No medicine is free of side effects, and the MMR has its own benefit/risk balance. Autism has never been one of those risks.)

As vaccination rates have slipped, measles has come back. In 2025 the United States recorded 2,288 cases, its worst year in more than three decades, and three people died: two unvaccinated children in Texas and an unvaccinated adult in New Mexico. They were the first measles deaths in this country in a decade. In 2026 the cases have kept coming at a similar pace.

Those are preventable via community vaccination. We’ve already shown it can be done. That is the real weight on the other side of the scale: a childhood disease we have already beaten is coming back to kill children.

Even the country’s most prominent vaccine skeptics have begun, grudgingly, to concede the point. Robert F. Kennedy Jr., now the nation’s health secretary, spent the better part of two decades helping to spread the very types of fears this essay is trying to undo. Yet standing in a Texas county where two unvaccinated children had just died of measles, he admitted that “the most effective way to prevent the spread of measles is the MMR vaccine,” even as he stopped short of urging anyone to get it. As the outbreak has worsened into 2026, the physicians he installed atop the health agencies have said it more directly, urging vaccination to protect whole communities.

Part of me wants to welcome this as progress (and I suppose part of me does). But I can’t quite feel it in my bones, because these admissions come so late and at such a price. When the people entrusted with protecting children spend years lending their credibility to a fear built on a fraud, a half-admission toward the evidence, offered only after children have already died, is a woefully thin kind of leadership. The children who catch a preventable disease in the meantime don’t get a do-over. That is why I believe that there is serious work to be done, and why I believe that careful and nuanced explanation still matters, and why I am sitting here at my desk at 5 am writing this while my family is still asleep (dog included).

No one wants to hear “vaccines are safe and effective” for the umpteenth time.

And yet, I do think that people have a real appetite for the why behind important health recommendations for our children (and for us).

If you want to support my work, a paid subscription is the best way to do it. It buys me the time to dig through primary sources and medical records, with no ads and no sponsors shaping a single word of what I write here. That independence is a direct result of your support, and I am grateful for it.

-Morgan

Morgan McSweeney, PhD (@dr.noc)

Part 1 of 3 on artificial sweeteners: a scientist reads the cancer scares and 50 years of toxicology, and what kept coming back null.

Morgan McSweeney, PhD (@dr.noc)

Every few months a new headline turns a sweetener into a suspect: aspartame is a “possible carcinogen”; this product is “linked to strokes”; the stuff in your coffee is “wrecking your gut.” Each one lands as a little jolt of dread and leaves you standing in the kitchen holding a can of Diet Coke wondering whether it is a live grenade.

I went and read the actual studies, the alarming ones and the reassuring ones, in animals and in people. My blunt read: the non-sugar sweeteners are some of the most heavily tested additives in the entire food supply, and the scariest claims about them (cancer, seizures, birth defects) keep coming back null in the studies built to catch them.

However, the opposite type of headline, the “it’s totally inert” line of messaging, also feels like an oversimplification that I have never felt particularly comforted by. So, I decided to look more closely.

I write these letters early. Lately, my research and writing routine is a very early morning session, starting around 4am, when everyone else in the house (including Franklin, our dog) is still asleep. But sometime around 5 or 5:30am, I usually hear Franklin let out a big yawn halfway across the living room. When I look over from my desk at the side of the room, I can see him doing his big stretch, and then he sleepily walks his way over to stand at my side with his tail slowly wagging as a greeting. That is his way of politely asking if he can go outside to go to the bathroom, which is my cue to give him a scratch on the chest, turn off the house alarm, and let him into our backyard. And, 99% of the time, I remember every step in that sequence: 1) scratch him on the chest, 2) turn off the alarm, 3) open the door to let him go pee.

Alas, on those fateful remaining 1% of mornings, I forget step #2.

This morning was one such morning. Everyone else, sleeping peacefully in the household, woke up to the ambience of the blaring alarm ripping through the house like a banshee who stubbed her toe. To add to the immediate chaos, right as the alarm started its pre-dawn shriek, Franklin happened to see the family of nine deer that live in the forest near our community, and he started barking like a terrified (but loyal) guard hound. So I had to run inside and make sure the alarm was turned off and then run back outside to assure our 55-pound retriever that the deer were no threat (despite their hoof-stomping). Zina sent me a worried “are you ok?” text, and I quickly responded “forgot the alarm while taking dog out.”

Now, here I am back at my desk, decidedly awake. Everyone has settled back to sleep after the disruption. Franklin is back inside, laying on the carpet next to me (not asleep, but also not anxious), and I am clicking through my browser tabs trying to remember where I left my train of thought.

A scary health headline does the same thing to your body that my house alarm did: a sudden blast of adrenaline that feels like danger. But, an alarm going off or a dog barking in the backyard only tells you that something moved. The only way to know whether it is a real intruder or just the same family of nine deer stomping their hooves in the tree line is to walk outside and look. That is the goal today: to take a look at the data on artificial sweeteners.

My takeaway is that, although I don’t personally drink soda in general, if I did drink sugar-sweetened sodas, I would be comfortable replacing them with artificially sweetened sodas.

What artificial sweeteners even are

There are six intensely sweet compounds the FDA has cleared as food additives: saccharin, aspartame, acesulfame-potassium, sucralose, neotame, and advantame. They are called “high-intensity” for a reason. Saccharin is a few hundred times sweeter than table sugar, sucralose about 600 times, and advantame is roughly 20,000 times sweeter, which is why the amount in your drink is measured in milligrams.

Most of them are not really “food” in the sense your body cares about. Saccharin and acesulfame-K pass through you and leave in your urine chemically unchanged. The majority of sucralose is never absorbed at all and exits in your stool. The one exception is aspartame, which your gut takes apart completely into two amino acids and a small amount of methanol. Importantly, all three of these metabolic products show up in ordinary foods and beverages (like fruit or milk), often in larger quantities than a diet soda delivers. As a note, that metabolic breakdown is also why aspartame carries a warning for people with phenylketonuria, a rare genetic condition in which one of those amino acids, phenylalanine, builds up dangerously.

Then, there are the “natural” sweeteners, stevia and monk fruit, that are marketed as the wholesome alternative. Upon my reading, the “natural versus artificial” line is a distinction without a difference: it is a marketing and sometimes a regulatory distinction, but it does not imply a different category of health risk. The stevia in your pantry is of course not a leaf; it is a purified isolate of specific molecules (steviol glycosides) because, ironically enough, the FDA has considered some of the less-purified, crude stevia extracts to have “inadequate toxicological information” (per Import Alert 45-06).

The WHO puts purified stevia in the same category as aspartame and sucralose and calls the whole group “non-sugar sweeteners.”

I prefer that approach rather than having two confusing groups of “artificial” and “natural” non-sugar sweeteners. The differences that actually separate these compounds are things like whether your body absorbs them and what your gut bacteria do with them, not whether a plant or a chemist made them first.

Sugar Alcohols (erythritol, xylitol, “-ol”)

One category gets swept into these arguments but does not belong there: the sugar alcohols, like erythritol and xylitol. Those are a different kind of molecule, are partly absorbed, partly caloric, are often responsible for the gastrointestinal complaints people sometimes blame on “sweeteners” in general. Those are different from the class of sweeteners you might find in Diet Coke.

They receive a lot of headlines, though, so let’s quickly cover some basics.

If you have heart disease or a history of blood clots, you may have heard some concerning news in the past couple of years. A few studies from one research group have tied higher blood levels of erythritol and xylitol to stickier platelets and to increased risk of heart attacks and strokes. I take that seriously, but the evidence is a lot shakier than the headlines suggest, for several reasons: it comes primarily from a single research group, it is observational data with potential confounding issues, and it turned up in high-risk patients already sick enough to be in a cardiac workup. The FDA reviewed the main erythritol study, discussed its merits and limitations in depth, and has not changed how it is regulated.

So, it is a real research signal, but it is one that nobody has yet shown to be a real-world harm, and that has not been replicated outside of the studies from that one research group. The data are simply not strong enough to warrant broad discouragement of sugar alcohols. And yet, the downside of easing off sugar alcohols is low, assuming you’re not replacing them with a bunch of added sugars. So, I can easily see how a reasonable person with a relevant medical history could decide to drop products with sugar alcohols.

(I will take more time with the sugar alcohols in a separate letter.)

What “possibly carcinogenic” actually means

In July 2023, the International Agency for Research on Cancer classified aspartame as Group 2B, “possibly carcinogenic to humans.” The phrase traveled far and wide over social media through misleading headlines like this.

IARC’s classifications describe how strong the evidence is that something can, under some circumstances, cause cancer. They deliberately say nothing about how much risk a substance poses at the doses people actually encounter.

Group 2B, in a case like aspartame’s, is the tier used when the human evidence is “limited,” meaning there is some suggestive evidence but it also could potentially be explained by chance or research method bias (the latter being more relevant in the observational studies for aspartame). It is a category that is shared by hundreds of agents including aloe vera extract and, believe it or not, pickled vegetables. Being in a category with pickles is not a reason to panic. On the same day IARC assigned that label, WHO’s food-safety committee reviewed the same evidence and reaffirmed the existing safe intake, stating there was “no convincing evidence” of harm from aspartame at the normal levels of consumption.

Two arms of the same organization, reading the same studies, and the shared interpretation between their findings is that the hazard may theoretically be on the table for the molecule, but the real-world risk at real-world intake levels has not been demonstrated.

Who has some reason to be extra cautious?

For most healthy adults, swapping sugar for a sweetener in moderation is a very defensible trade, as we will discuss in detail further below. But a few groups have reasons to consider limiting or avoiding.

• If you have PKU (phenylketonuria): don’t eat/drink aspartame. This is a hard rule. Aspartame breaks down into phenylalanine, which you cannot clear, which is exactly what that warning label on the can is for.

• If you are pregnant: the signal on artificial sweeteners is soft and not particularly worrisome, and the data probably say more about heavy soda drinking in general than about any specific sweetener, since several of the same studies often flag risk associations with sugary soda, too. However, this is another reasonable circumstance to ease off soda (of any kind) and switch to trying out some unsweetened beverages. But the specific choice of a replacement is important. I would support switching from diet soda to something like sparkling water with a slice of cucumber and lemon all day long. However, for example, the data do not support switching from diet sodas to sugary sodas in pregnancy.

For most people, these remain approved as safe within normal consumption levels, and in moderation they are a reasonable step down when used as a replacement for added sugars.

In the rest of today’s deep dive:

• How much do you have to eat/drink for it to be considered “a lot?” per regulatory standards

• The toxicology: the animal studies, the DNA-damage tests, and the largest human study ever run on this (roughly 474,000 people).

• One lab keeps finding cancer in animal models, and why regulators on both sides of the Atlantic set its work aside.

• The original sweetener cancer scare, saccharin and the rats, and the biological reason the argument fell apart so conclusively that the government removed the warning label from cans.

• “Doesn’t aspartame turn into formaldehyde?” A tiny sliver of truth that is greatly misrepresented. The embalming-fluid claim vs. a glass of fruit juice.

• Heart, blood sugar, seizure, headache, allergy, and pregnancy questions, and the available data.

• (Opinion) My bottom line: what I do about all of this.

A paid subscription also opens the full archive. A few reader favorites: i) The Heart Attack That “Came Out of Nowhere”, ii) The Truth: Is Alzheimer’s Preventable?, and iii) Creatine: What Works and What’s Hype.

This letter is not sponsored, there are no ads, and there are no affiliate links. It is entirely reader-supported, which I believe to be the best model to avoid commercial bias on coverage of important topics in health.

Subscribe now

If it is not useful, unsubscribe any time in one click.

Read more

What actually lowers your risk of dementia, what is a waste of your money, and the one minor exception.

Morgan McSweeney, PhD (@dr.noc)

Alzheimer’s dementia is the disease many people fear the most, and it is also one of the most heavily marketed, which I find to be a rather cruel pairing: there is a mountain of very expensive false hope aimed at a very real fear.

So I will lead with the optimistic result from my reading of the literature in preparation for this letter. A meaningful share of your risk (perhaps as high as 45%) is actually yours to move, the things that move it are unglamorous and mostly free, and the clock on them starts far earlier than you would guess, literally decades before you have a single symptom. Almost everything sold to you in a bottle for “memory” or “brain support,” on the other hand, does nothing.

Today, we discuss the critical details on both sides of that reality.

Our daughter recently got a playdough set that comes with little “mix in” toys. The theme was a fairy garden, so there was a fairy, a fairy cottage, a hedgehog, and a variety of mushrooms, including the classic red ones with the little white spots on top. Those red mushrooms were her favorite, and she asked “What is this one called, Daddy?”

I had no idea, so I told her we could look it up together. We learned that they are called Amanita muscaria, and now my two-year-old happens to know the scientific name for what I previously would have just called “red fairytale mushrooms.”

The funny part is that those mushrooms turn up surprisingly often once they are top of mind, and, right now, they are very top of mind in our house. We see them everywhere: at the science museum, on kids’ backpacks, in pieces of art.

Every single time, she points and announces, with total delight, “Amanita muscaria!!” And, if they are in conversation with us, the nearby adults look completely baffled. Not because she hasn’t said the name clearly, but because they don’t know the scientific name for the red fairytale mushrooms themselves. Or, I suppose there is a chance they do, and it is just so far down the list of things they expect a two-year-old to say that their brain says “that can’t possibly be what I just heard.”

Watching a brand-new mind grab an eight-syllable Latin name and hold onto it for weeks is one of my favorite things about this stage of toddlerhood. I am repeatedly impressed at how eager-to-learn she is. It also, if I’m being completely honest, makes me think about the other end of that curve, and about me staying sharp enough to keep up with her mushroom identification skills for as long as I possibly can. Which brings me to the topic of today’s deep dive:

“Is it true that you can prevent Alzheimer’s?”

Here is my answer after a review of the literature, and I’ll spend the rest of this letter backing it up in detail. “Prevent” has the right spirit but is ultimately the wrong word. For the common, late-onset Alzheimer’s that most people picture and fear, there is no intervention that guarantees you won’t get it. But “no guarantee” is not the same as “nothing you can do.” The largest analysis in the field estimates that up to about 45% of dementia cases could, in theory, be prevented or delayed by factors you can actually influence. Treat that 45% as a ceiling rather than a promise, but the direction is not in doubt: a real, meaningful chunk of this is modifiable.

And a quick word on age, in two directions. If you are in your 30s or 40s and think of this as “a problem for my 60s,” don’t. The factors that carry the most weight are midlife factors, your midlife numbers are being set right now, and the disease process itself is already underway in your blood vessels and your brain long before any symptom emerges, often decades ahead of a diagnosis. And if you are in your 50s, 60s, 70s, or 80s, the opposite worry, that you have already missed your window, is just as wrong. The Lancet Commission’s own conclusion is that it is never too early and never too late to lower your risk, and several of the biggest risk-reduction factors have been clearly shown to pay off when you start later in life. So, wherever you are on that curve, you need to find out which factors actually shape your risk, and start working on them.

If you read between the lines, you should note you can do everything “right” and still end up with dementia, and a diagnosis of dementia is therefore not to be regretted as a failure to have done “everything you could” to reduce risk. Like most other topics in health, this is a question tilting the odds in our favor, nothing more.

But the scientific reality is that the things that actually move risk downward are mostly boring and mostly free, and almost everything else being marketed to you for “brain health” and “memory support” is worthless.

Let me start by telling you a few things to stop paying for.

If you have seen the ads for Prevagen, the jellyfish-protein memory pill, you should know that in 2024 a jury found its advertising deceptive and fraudulent. In the company’s own study, the one they built the ads on, Prevagen failed to beat placebo on all nine memory tasks they measured. So, the researchers sliced the data into more than thirty subgroups until something (anything) turned up “significant” by chance. And apoaequorin, the active ingredient, is a protein, which means your stomach digests it into amino acids like any other bit of food, and no significant fraction could even possibly reach your brain. Yikes. It’s a placebo jellyfish-protein pill.

And yet, they will happily charge you $60/month for their “extra strength” formulation.

Save your money.

“Brain training” games? Lumosity paid a $2 million FTC settlement for claiming its games could stave off dementia and Alzheimer’s, which it could not support with any evidence.

“The creators and marketers of the Lumosity “brain training” program have agreed to settle Federal Trade Commission charges alleging that they deceived consumers with unfounded claims that Lumosity games can help users perform better at work and in school, and reduce or delay cognitive impairment associated with age and other serious health conditions.”

I guess you could play their games if you find them fun, but if you think you are paying for dementia risk reduction, spend your $12 per month elsewhere.

Ginkgo biloba? To its credit, there was a randomized controlled trial. But that trial followed more than 3,000 older adults for six years and found no reduction in dementia at all. “The overall dementia rate was 3.3 per 100 person-years in participants assigned to G. biloba and 2.9 per 100 person-years in the placebo group.”

Coconut oil? One viral anecdote and no clinical trials. No convincing evidence.

“Until we have the results from a large-scale clinical trial, we must conclude that at the moment there’s no scientific evidence that coconut oil can help people with dementia.”

Two big takeaways even if you read no further:

• The biggest real controls on your Alzheimer’s risk are simple and mostly free: your blood pressure, your cholesterol, correcting hearing loss, how much you move each day, your midlife metabolic health, and more we will discuss. None of them come in a bottle, which is exactly why most people overlook them and spend attention and money on the products instead. Below, I go through each one with the actual trial evidence, how big the effect really is (a few are smaller than the headlines suggest), and specific things to aim for.

• Nearly everything sold to you in a bottle for “memory” has either never been properly tested or has already failed a trial. There is exactly one exception with real, if modest, evidence, and it is a plain, cheap multivitamin, not the fancy “brain” pills. But there is an important catch with that study that changes my perspective on whether you should actually bother, and I cover it below.

You should note that the largest drivers of risk for dementia (bullet #1) overlap with many of the key drivers of every chronic disease from metabolic disease, to heart disease, and even some types of cancers.

Here is everything in the rest of today’s letter on how to protect your brain, with every study linked so you can check my work:

• The realistic goal to aim for, and why “prevention” is the wrong target

• The factors that actually work, their trial data on how much they each lower your risk, and the findings that are shakier than the headlines admit

• The “Third Space” move: one simple habit that targets several risk factors at once

• Save your money, part two: more high-profile “memory” products to not buy, and the studies that sank them

• The multivitamin data, and the catch

• A closer look at the data on what the new Alzheimer’s drugs can and can’t do

• (Opinion) What I personally do with this info, and how I prioritize

Who this is for: anyone worried about dementia who wants a clear description of the evidence on risk reduction… whether you have watched a parent or grandparent go through dementia, or are getting older and want to know what has been shown to reduce risk, or are younger and want to get a head start on the decades-long process of risk reduction.

You will come away with a better understanding of what to prioritize vs. what is a waste of your money.

I run no ads, take no sponsorships, and use no affiliate links. On a topic this crowded with people selling you false hope, that independence is the whole point. I would much rather you pay me directly for my time spent reading and translating the evidence, rather than to take money as a cut of sales from some product I recommend you buy.

A paid subscription also opens the archive and gives access to every deep-dive yet to come. If you’re ready to check out some prior letters, two popular ones are: i) The Heart Attack That “Came Out of Nowhere,” and ii) Creatine: What Works and What’s Hype. Upcoming deep dives will cover artificial sweeteners, the gut microbiome, probiotics / prebiotics, and more.

The monthly option is $12/month and the annual option works out to $8/month, which means about $1 per deep dive letter.

Subscribe now

If you do not find the letter to be useful, you can unsubscribe any time in one click.

Now, let’s talk about what you can start doing today to protect your brain in the long term.

Read more

-Morgan McSweeney, PhD (@dr.noc)

A week ago I published a letter on the parasite called Cyclospora that has been speading across the country while the CDC could barely count it: their official tally at the time said 145 cases, and state health departments were already reporting nearly ten times that.

Important update from yesterday: The CDC has now linked a major outbreak to the shredded iceberg lettuce served at Taco Bell, leading to more than 1,644 people sick and 94 hospitalized across 5 states (Indiana, Kentucky, Michigan, Ohio, and West Virginia). The FDA traced those Taco Bell-linked cases to a single lettuce grower in Mexico.

But Taco Bell is very unlikely to be the whole story. The CDC’s own notice says it is “also investigating other outbreaks and illnesses of cyclosporiasis nationally that are unrelated to this outbreak.” Explicitly, that means that it is not just from Taco Bell.

Their lettuce is just the tip of the… iceberg

The NYT’s reporting backs that up: Dr. Natasha Bagdasarian, Michigan’s chief medical executive, noted that not everyone who got sick had eaten at a single common restaurant, and many reportedly never ate at Taco Bell at all. At peak season (now) this parasite can ride in on several different products at once.

It is an open question how far that grower's lettuce spread beyond Taco Bell, and how much of the wider surge in the country is coming from that same lettuce (sold through other venues), vs. from entirely separate products.

2026 is on pace to be the worst Cyclospora year on record, with thousands more illnesses still under investigation, right as a lot of state and local health labs face budget cuts and staff layoffs.

Not a great time to be decimating public health infrastructure.

A few things the generic “rinse your produce” reminders skip:

This is a parasite, not a 24-hour bug. Symptoms (watery diarrhea, no appetite, fatigue, cramping) usually show up about a week after exposure and, untreated, can drag on for a month or more, often in a relapsing pattern where you feel better and then it comes roaring back. It can eventually clear on its own, but the reliable way to shorten it is a specific antibiotic your doctor can prescribe. I wrote a more detailed description of the data supporting trimethoprim-sulfamethoxazole for Cyclospora in my prior letter from July 10th. (See the section The cure has pretty good data, and a placebo-controlled trial is why).

You can’t catch it directly from another person. Unlike a stomach bug, Cyclospora isn’t contagious. The parasite has to sit out in the environment for one to two weeks before it can infect anyone, so it only spreads through contaminated food or water. You won’t catch it directly from a sick family member, a coworker, or whoever assembled your lunch.

Your doctor may not test for it unless you ask. A routine stool test often doesn’t screen for Cyclospora, so if you’re sick and you’ve eaten suspect produce, say the word “Cyclospora” out loud to your provider, and make sure whatever stool test they run actually includes it as a target, because not all of them do. Your doctor may have a good reason why they do not think Cyclospora is a likely reason for your diarrhea (which could be due to many other causes), but it’s a good idea to make sure it’s been considered if your symptoms seem to drag on and relapse.

As I discussed in more detail previously, rinsing may help but doesn’t reliably remove it. Only cooking to about 158°F is expected to kill the parasite, which doesn’t help much for a salad or lettuce on a taco.

Subscribe now

(OPINION WARNING).

The below is my own take, as a scientist and a parent, since people keep asking me what I, specifically, am doing

“Would you pull all fresh produce out of the house until this is over?”

Me? Absolutely not.

Cutting fresh fruits and vegetables out causes a small but certain harm every single day. One of the best things I do for my body on a regular basis is eat a lot of whole, plant-based foods, which for me includes a lot of fresh berries, avocados, and so on. And, I believe that my odds of catching Cyclospora are very low (as is true for most people).

So, especially for families without specific health circumstances that warrant extra caution, I feel that cutting out fresh fruits and vegetables because it’s a bad Cyclospora season would be similar to cutting out going to the gym because it’s a bad cold and flu season.

I can’t very well put a number on the tradeoff. Of course, if you do get sick, a week or two of miserable diarrhea is much more harmful to your health than the nutrition you would have lost from sacrificing fresh produce for a few weeks. But, you have to multiply that bad outcome by its probability. Without good estimates of the relevant numbers, we simply can’t do that calculation. My feeling – separate from any data - is that totally giving up fresh produce for weeks-to-months does not come out ahead for our family.

So instead of quitting fresh produce, we have simply adjusted the mix. We’re having more smooth- and thick-skinned or peeled options (blueberries, citrus, avocado, watermelon) over the crevice-heavy berries and bagged salads, and we are just skipping lettuce for now.

(If you choose to still eat lettuce, a whole head you wash yourself beats a “pre-washed” bag, and re-rinsing that bag won’t remove the parasite and can even add contamination from your sink).

Produce from your own garden is fine (unless people have been poopin’ in your garden.)

One thing I would advise you skip: the veggie-wash sprays and vinegar or bleach soaks, which do not kill Cyclospora and may give you a false sense of security. I showed some of those data last time.

If someone in your home is pregnant, elderly, immunocompromised, or very young, that is a good reason for more caution, like opting for more cooked produce, postponing lettuce, etc.

Stay hydrated, ask your doctor about whether you should specifically test for Cyclospora if you’re sick, and don’t eat lettuce from Taco Bell at the moment in Indiana, Kentucky, Michigan, Ohio, or West Virginia.

-Morgan

Morgan McSweeney, PhD (@dr.noc)

I’m a PhD scientist and science writer, not a physician; this is general education, not medical advice

A scientist reads the actual trials: the inflammation claim is weak in human studies, the disease link points more convincingly to ultraprocessed food at large, and even RFK Jr.’s own MAHA report didn’t call seed oils harmful.

Morgan McSweeney, PhD

Walk down any grocery aisle in 2026 and you’ll see it: “NO SEED OILS,” stamped on chips and salad dressing like a warning label. Restaurants are switching their fryers back to beef tallow (and bragging about it). And the Secretary of Health and Human Services, Robert F. Kennedy Jr., has called seed oils “one of the most unhealthy ingredients that we have in foods” and says Americans are being “unknowingly poisoned” by them.

I’m a scientist, and I’ve watched this scare build for a couple of years. Here’s my blunt read on the evidence: the case against seed oils is one of the weakest big health scares I’ve come across, and the most prominent claims about it don’t survive thorough cross examination against the data.

When the “Make America Healthy Again” commission published its flagship report in May 2025, its discussion of seed oils was strikingly more mild than I anticipated. The report noted that refining these oils strips out some micronutrients like vitamin E, and that the oils “contribute to an imbalanced omega-6/omega-3 ratio, a topic of ongoing research for its potential role in inflammation.”

That is as far as it went: it’s a topic of ongoing research, a potential role, with no “poison” claims and no call to ban anything. It is a careful, hedged type of claim, and it contradicts the Secretary’s own “poisoning you” language and podcast bravado.

When the report written by the people leading the charge against seed oils won’t even call seed oils harmful, that’s a clue.

The mainstream scare is founded upon one main correlation

Here’s a core argument you’ll see across podcasts, facebook, twitter, and elsewhere: seed-oil consumption has soared over the last century, and, over the same timespan, obesity and chronic disease have soared too. So, seed oils must be behind it.

The trouble is that a lot of other things rose over that same century. And one of them rose right alongside obesity and has a mountain of evidence behind it: ultraprocessed food. Ultraprocessed food now makes up 54% of the calories Americans eat at home, up from 51% two decades ago, and increased consumption of it is reliably linked to worse health across dozens of outcomes, from heart disease to early death.

Now, ask where most people actually get their seed oils. Not from a spoonful of canola while cooking at the stove. They get seed oils most commonly inside ultraprocessed food.

So, are the negative health effects of ultraprocessed foods driven by their inclusion of seed oils? Or, is it another factor (or combination of other factors), such as ultraprocessed foods’ typically high levels of added sugars and salt, their lack of dietary fiber and other nutrients, and their high energy density combined with hyperpalatibility.

Although there is nuance to this discussion, I think the latter is more likely, and you will soon see why.

“But seed oils cause inflammation”

This is the scientific heart of the fear, and it rests on a misunderstanding.

Seed oils are high in omega-6 fats, mainly one called linoleic acid. Omega-6 and omega-3 are both essential fats, meaning your body can not make them and has to get them from food. Somewhere in the past few years, the internet decided omega-6 was the evil twin of omega-3: if omega-3 is anti-inflammatory, then omega-6 must do the opposite. That is simple and attractive as a mental model, but not quite biologically accurate. Omega-6 does many of the same jobs omega-3 does, and one of the things linoleic acid (the omega-6) does is lower your LDL cholesterol, which is a good thing.

More important than the mechanistic discussion is the human data: when researchers stop theorizing and actually feed people more linoleic acid, then measure their inflammation, the inflammatory markers don’t go up. That systematic review of the randomized human trials concluded:

“virtually no evidence is available from randomized, controlled intervention studies among healthy, noninfant human beings to show that addition of LA to the diet increases the concentration of inflammatory markers.”

So the “seed oils are inflammatory” claim is a lab hypothesis that sounds scary on paper (or in an Instagram reel), but that simply has not proven to be true in real people.

Follow the money

So who benefits from the seed oil scare? Start with the label itself. “Seed-oil-free” has become a booming marketing claim: sales of certified seed-oil-free products jumped 216% in a single year. Steak ’n Shake made national news switching its fries to beef tallow, and even had the Secretary of Health in for a bizarre promotional spot with Fox News’s Sean Hannity. Because I suppose french fries and burgers are MAHA-approved as long as they are prepared using beef tallow rather than vegetable oil?

One frozen-food company that changed its product line said it chose beef tallow partly for its “lower price point.” Companies are taking advantage of confusion around the health effects of seed oils to switch to cheaper sources of fats that are actually probably worse for your health (more on that below).

But yes, the money runs both ways. The seed-oil and processed-food industries are enormous, and they have their own lobbies and their own history of funding friendly nutrition research. That’s why I’m anchoring this on randomized trials and not on the government’s MAHA report or the industry’s own report.

So, where does that leave us, and what should you do?

The scare is built on a correlation that fits ultraprocessed food far better than it fits the oil, and its central claimed mechanism (inflammation) simply doesn’t show up when you actually test it in people and measure for inflammatory outcomes.

But the most important question is bigger than “are seed oils poison?”

Does swapping animal fat out for a seed oil actually do anything good for your heart, and does it increase or decrease risk of death?

That’s what the human trials I walk through below were built to answer, including the two that seed-oil skeptics most commonly cite. I’ll lay it all out, limits and all, including the one place where the skeptics actually have a good point.

Because “seed oils are not the villain” is not the same as “eat whatever you want.” The most useful question was never whether they’re poison; it’s what to actually put on your plate, and that’s where the evidence below turns practical.

Below, the full detail for paid subscribers:

• The best-controlled experiment in the entire seed oil debate, and the reason its result actually runs counter to the panic.

• Two trials seed-oil skeptics commonly cite, and what they actually say once you read past the headline.

• What swapping animal fat for a seed oil does to your cholesterol, your heart, and your lifespan.

• Why the omega-6-inflammation mechanism falls apart if you trace it step by step.

• And the part you can actually use: what the best evidence says to do about your diet, why “cut it out” is the wrong instinct, and the almost-too-easy way to add the one thing (fiber) that is more important for your health than any oil in your cupboard.

Paid subscribers get the fully sourced deep dive behind the health headlines, twice a week, for less than a dollar a letter. One note on timing, since it’s relevant this week: the rate goes up this Tuesday July 21st, from $8 a month (or $80 for the year) to $12 a month (or $96 a year). Subscribe before then and Substack locks you in at today’s rate for as long as you stay subscribed, even after the increase. So, if you’ve been thinking about it, this is the week.

Subscribe now

I take no ads, no sponsorships, and there’s nothing in your cart at the end. That independence is the point, which is the whole reason my readers, not advertisers, keep this letter going. It’s the reason I can tell you if a popular scare is overblown, even when fanning the flames of uncertainty would do a better job at driving clicks and attention.

Let’s get into it, starting with an important experiment on seed oils and cholesterol.

Read more