Hi {{first name | there}},

A quick welcome to everyone who joined from YouTube this week, whether you came for the Heart-Test Map or the Supplement Stack. Both guides are linked at the end of this note.

Creatine gives me insomnia.

I still recommend it.

This basically sums up supplement advice, and much of medicine.

The evidence can be good. The intervention can still be wrong for any one person.

For personal and professional reasons, I have tried dozens of supplements myself and recommended many to others.

New evidence changed my guidance. So did results I've seen in myself and my patients.

Today, I want to show you three examples demonstrating the nuance of supplement stacks. And this is just supplements!

I once treated 40 to 60 ng/mL as an optimization target and K2 as the automatic partner to D3. Then large randomized trials found no broad benefit for cardiovascular disease, cancer incidence, fractures, or depression in generally healthy adults. I swung toward “mostly useless.”

Both extremes were over-generalizations.

Vitamin D is essential. It helps the body absorb calcium and maintain bone. The mistake was treating that fact as proof that pushing an already adequate level higher would prevent a long list of diseases.

VITAL randomized more than 25,000 adults to 2,000 IU of D3 a day or placebo. It did not reduce the main cancer or cardiovascular outcomes. But among participants with measurements, the average starting level was 30.8 ng/mL. The trial mostly showed that giving more vitamin D to people who already had enough did not make them healthier. But these results wouldn't apply for someone with a level of 8 or 10 ng/mL.

I now think of vitamin D, as I do many other vitamins, as a threshold nutrient.

Too little matters. Once you have enough however, more usually does not help. A result of 8, 28, and 78 ng/mL should not lead to the same supplementation plan.

In my most enthusiastic days, my 25-hydroxyvitamin D level reached 70 ng/mL while I was taking 5,000 IU of D3 plus 100 mcg of K2. While this doesn't count as toxicity (yet), it was probably a waste of effort and money.

K2 is another interesting case.

I used to tell people emphatically to pair D3 with K2, especially at higher doses. The mechanism makes sense: vitamin D increases calcium absorption, while vitamin K activates proteins involved in bone and vascular tissue. The popular shorthand says K2 “directs” calcium into bone and away from arteries.

But then I scoured the literature.

In AVADEC, 365 older men with aortic-valve calcification took 720 mcg of MK-7 plus 1,000 IU of D3 or placebo for two years. The vitamin K levels improved. Aortic-valve and coronary calcification did not slow overall.

I now consider K2 optional for most healthy adults taking ordinary D3, not a required companion and not protection against an excessive D dose. K2 can also interfere with warfarin and other vitamin K antagonists.

I still like vitamin D. I like it for correcting a real deficiency or covering predictably low input.

I no longer use a high number as the goal, and I no longer add K2 by default.

Creatine is one of the few supplements I can discuss without starting with cell or animal data.

There's real evidence showing that creatine helps muscles recycle energy during short, hard efforts. In practice, that may mean one more good rep. Over time, those small additions can matter. A meta-analysis of 22 trials in older adults found that creatine added to resistance training produced about 1.37 kilograms more lean tissue, plus small strength gains, compared with training alone.

Creatine did not replace training. It added a modest benefit to training. If someone does not lift, creatine won't help one iota with building strength or muscle.

Luckily, creatine is also pretty simple. You just need the cheap version: creatine monohydrate, usually 3 to 5 grams a day. No premium form has proved that it changes the outcome. Loading is optional and may cause more discomfort at first (GI disasters!). Cycling is unnecessary.

Then there is my own result.

When I take creatine, my sleep maintenance gets worse. I have tried it more than once, and have observed this in other individuals. And before you mention it, moving the dose earlier in the day does not solve the problem, which makes sense because creatine builds up in tissue rather than acting like a quick stimulant.

I cannot tell you for sure the mechanism behind my response. I have theories it has to do with improving brain energy (ATP) at the wrong time and place in an already rested brain. In any case, the results have been clear enough for me to stop.

Lying awake at 3 a.m., no meta-analysis will convince me to push through :D (I made a 54-second video about this, and the comments are full of people with the same story.)

This is where supplement advice often becomes strange. We want one verdict: good or bad, works or does not work. But the real decision weighs pros vs. cons for a specific individual's context (genes, lifestyle, etc.).

Ask first, does the product help similar people with this goal on average? Second, is the trade worth it for this person?

For me, a modest training benefit is not worth worse sleep. For someone who lifts, tolerates it, and wants a small edge in strength or lean mass, creatine remains a reasonable and affordable choice.

Beyond my own example, there are other tradeoffs too. Early water-weight gain and stomach upset are common reasons people stop using creatine. Creatine can also raise serum creatinine without reducing true kidney filtration, which can make a creatinine-based eGFR look worse (not a super rational reason, I'd argue, but this does stop quite a few folks). I've had patients also just stop because it made them feel "sloppy" due to the water retension.

Nicotinamide riboside, or NR, and nicotinamide mononucleotide, or NMN, are a harder sell.

The theoretical mechanism is not the problem. NR and NMN are precursors the body can use to make NAD, which cells need for energy metabolism and DNA repair. But the claim that NAD simply falls as we age is less settled in humans than it sounds. A 2026 analysis across seven human cohorts found that whole-blood NAD stayed stable with age. That does not settle what happens inside every tissue. Oral NR and NMN can still raise NAD or related metabolites in blood. I understand why this category took off.

I tried NR myself and did not notice a benefit.

My experience proves very little, and you and all your friends shouldn't make a decision based on it. But if the goal is more energy and you feel no difference, we do not have strong evidence for a hidden longevity payoff that should persuade you to keep taking it.

There is some evidence, albeit weak, supporting use of NAD+ and its precursors. In 25 postmenopausal women with prediabetes, NMN improved muscle insulin sensitivity. In 90 people with peripheral artery disease, NR improved six-minute walking distance. Those are useful leads, but tiny studies. They do not prove that a healthy adult has an NAD deficiency, will feel more energetic, avoid disease, or live longer by raising a blood NAD value.

I also have patients using subcutaneous NAD+ who insist it gives them a real energy boost. They have not reported adverse effects to me, and I take that seriously. A benefit someone feels repeatedly is hard to ignore, even if there is a chance of a placebo effect.

Fwiw, this route has barely been studied. A 2026 preliminary safety report included only six people who received subcutaneous NAD+ for three days. It was not peer-reviewed and did not test energy or long-term outcomes. Because these are compounded injections, sourcing and sterility matter.

A short-term energy boost is not a longevity outcome. But it certainly may be worth the tradeoff for someone who consistently feels better, while doing nothing to slow aging or extend life.

The broader trial record remains underwhelming. A 2025 meta-analysis of 12 NMN trials found higher blood NAD without consistent improvement in glucose or lipid measures. Another analysis of NMN and NR trials found no overall improvement in muscle mass, grip strength, or gait speed in older adults.

Taken together, the studies do not support a general benefit for healthy adults.

If your goal is to raise blood NAD, oral NR or NMN may do that. If your goal is reliable energy, better function, or a longer healthy life, the evidence is not there for generally healthy adults.

Before I add anything to my supplement stack, I try to complete one sentence:

I am taking this because ___, and I will know it is helping when ___ by ___.

Then I ask four questions:

Vitamin D passes when it corrects a real gap; K2 is no longer automatic. Creatine passes on the evidence and fails on my personal tradeoff. NR and/or NMN raises the marker, but the outcome I care about remains unclear.

But the full supplement landscape is near infinite, which is why I use this framework.

If I cannot answer the sentence, I do not have a good reason for adding this supplement (or any intervention!). Instead, I have a hope with a recurring expense.

I wanted a place to apply that test across the products people ask me about most, without pretending one stack fits everyone. So I rebuilt The Longevity Supplement Stack, the original of which you got upon subscription (but now much improved). It covers what has a clear job, what only makes sense in context, what remains interesting but unproven, and what I would skip first.

Get The Longevity Supplement Stack

Keep learning,

Hillary Lin, MD

P.S. Know someone whose supplement stack keeps growing? Send them the free guide: The Longevity Supplement Stack

P.P.S. If your cholesterol labs and heart scans seem to disagree, my free Heart-Test Map explains what ApoB, Lp(a), CAC, CCTA, and stress tests can and cannot tell you.

Today's three cases are one method applied three times. This episode is the whole method in 40 minutes.

Biohack Smarter: How to Test What Actually Works

How to take a plausible claim, turn it into a safe personal test, know what to measure, and know when to stop.

Watch on YouTube | Subscribe to The Longevity Show

☀️ Ten times the vitamin D, zero extra survival. In 455 patients with metastatic colorectal cancer, 4,000 IU a day beat 400 IU on nothing that mattered. The threshold idea held up in the hardest setting there is.

💉 A supplement powder is not an injectable. The FDA tied compounded glutathione made from dietary-supplement-grade ingredients to at least 30 adverse events, some shock-like, some ending in the hospital. Ask your compounder what grade they use.

🎯 Pancreatic cancer finally has a targeted pill. The FDA approved daraxonrasib, the first drug to hit the RAS mutations behind most of these tumors, for metastatic disease after prior treatment.

🩸 One sensor, two numbers. The FDA cleared Libre Duo, the first wearable that tracks glucose and ketones together, so people with diabetes get a warning before ketoacidosis becomes an ER visit.

🫀 Hit the LDL target twice as often. Same number of heart attacks. Starting evolocumab right after a heart attack got 82% of patients to goal versus 40%, and one-year events did not budge. A biomarker win is not an outcome win. Sound familiar?

🤖 A robot will take your blood now. The FDA authorized Aletta to find the vein, place the needle, swap tubes, and bandage you, with a human watching. Bad-stick veterans, this one is for you.

Science of Skin Summit, September 17-20, Austin, TX. I am speaking on AI in dermatology and on skin and hair as windows into biological age.

MVMNT Longevity Summit, September 22-23, Coronado, CA. I am leading a session on turning longevity data into clinical action.

Livelong Women's Health Summit, September 25-26, Hilton Midtown, New York City. I am moderating four panels and joining the closing audience Q&A.

Science of Skin Longevity Summit, February 19-21, 2027, Scottsdale, AZ. I am joining the faculty for a practical skin-longevity talk and the peptides panel.

Partner links support The Longevity Letter. They do not change the evidence grades in the guide, and none of these products belongs in every routine.

For physician-guided biological age testing, start on the testing page. Options include epigenetic age testing with a consultation, as well as organ-specific proteomic and SystemAge assessments when they fit the question.

Have a question or topic you want covered? Reply to this email.

Was this forwarded to you? Subscribe here to get The Longevity Letter in your inbox.