Discernment Dispatch: July 2026
Three pieces of research, two questions, one discernment practice to carry forward
Every month, I take three of the health claims making noise—the ones surfacing in your feed, your group chats, and your inbox—and hold each one up to two questions: Whose evidence is actually shaping this? And who benefits if I follow it? The aim isn’t a verdict. It’s to show the questions working, so they become your own questions too—questions you reach for the next time a promise arrives already sounding like settled science or your next solution.
You can read these three as reflections, not rulings.
What matters isn’t where I land on any one of them. It’s the habit you carry past them, so the next claim meets you from a more informed, more discerning place. This is how we relearn discernment and regain true empowerment in our own health.
Remember: Everything is connected. We are all unique. And all things matter.
This is a sample of the Discernment Dispatch—the monthly members-only edition I publish at the end of each month, exclusively for the Inner Circle. Inner Circle members receive the Dispatch on top of everything in the subscriber tiers below it: the weekly Chat and all live gatherings. Most months, the Dispatch is for those members only. I’m sharing one in full so you can see if it’s the kind of thinking and conversation you want more of. I’d love to have you join me in the Inner Circle.
1. An allergy pill and a heartburn pill are the new fix for hot flashes.
This pairing is everywhere this month: Allegra and Pepcid—fexofenadine and famotidine—taken together for hot flashes, brain fog, the three a.m. sweaty wake-up. Thousands of posts with women describing real relief.
The respite isn’t imaginary. But hold the enthusiasm before deciding anything.
Whose evidence is shaping this? The trend didn’t come from a trial. It spread through the online grapevine, on the strength of testimony.
The relief itself has two potential explanations:
Both drugs block histamine—the same compound behind allergy symptoms. The theory is simple enough: histamine seems to play a role in hot flashes, so blunting it might quiet them. Plausible on its face, but the mechanism is shakier than it sounds. Histamine does many jobs in the body—including helping regulate temperature—but its role in hot flashes isn’t actually settled. No one has tested this combination in people, for this purpose. We don’t know if it works, and we don’t know what daily use does to a body over years. (Blocking histamine has a price.)
Then there’s placebo. And hot flashes produce one of the largest placebo effects in medicine. Count it by women, and 30 to 50 percent feel real relief on a sugar pill. Count it by the effect, and in trials of paroxetine—the first FDA-approved non-hormonal option for hot flashes—placebo accounted for nearly 80 percent of the benefit, leaving barely a fifth to the drug itself.
Feeling better about a remedy and the remedy working are easy to mistake for each other. The bodies feeling better are telling the truth. The remedy attributed to that feeling is the unverified part.
Who benefits? The person posting does. The share, the save, the algorithm rewarding the woman whose video reaches the next woman and whose message spreads like wildfire.
The body taking two daily drugs for the long haul carries a cost no caption mentions. Long-term acid suppression—famotidine’s entire mechanism—can interfere with how the body absorbs key nutrients like B12, magnesium, iron, and calcium, the reserves a menopausal woman is already working to protect for her bones and her blood. The pills arrived where curiosity would have served better.
Off-label use of a pharmaceutical drug is the prescribing or use of an FDA-approved medication for a purpose, population, dose, or route of administration that falls outside the specific parameters that have been evaluated and approved for that product. The term signals only that the use has not been reviewed by the FDA for the claim being made. It doesn’t mean the use is unsafe or ineffective. It also doesn’t mean it’s supported by evidence. Some off-label uses have decades of research behind them. Others have none.
The pharmaceutical industry stands to benefit too, though not in the obvious way. Fexofenadine and famotidine are decades off-patent, which is why they’re cost-effective on the shelf. The generic sales aren’t where the money would be. The opportunity is downstream. When an off-label use scales to this size, it becomes a market signal: proof of demand for a menopause symptom that a company could meet with a patentable drug of its own. The free, accidental trend becomes the market research for the paid, deliberate product that follows.
How to hold it: The relief may be real. But the explanation is unproven and could have downstream consequences. Both are true at the same time. The protocol may work to alleviate the symptom, but it’s not worth starting simply because a stranger’s video gave us permission without considering that the impacts outside of the relief could be significant. That same influencer or friend isn’t there when the nutrient deficiencies surface. They’re just there for the hot flash, the complaint, and the band-aid.
A hot flash is information. Silenced before it’s read, it becomes a clue no one followed. Silenced by drugs that interrupt the body’s signaling without addressing what the signal pointed to, you may be trading the symptom for what the drugs themselves do over time—a different problem, on a longer timeline.
2. Creatine is the supplement midlife women suddenly can’t do without.
It’s the canister that showed up on the counter this year, a scoop stirred into the morning coffee. Once a bodybuilder’s powder, creatine is now sold to women in their forties and fifties for muscle—but also for brain fog, mood, sleep, and the flurry of midlife symptoms. It may work, at least for the muscles. Creatine is among the cheapest and most-studied supplements there is. This is not a debunk of that particular fact.
Whose evidence is shaping this? For strength and muscle, there’s decades of solid research on creatine. For the brain-and-mood promise actually driving the trend, it’s something far flimsier—a handful of small, short studies, one eight-week cognition trial in just 36 women, using a premium formulation of creatine. Even that small trial showed a reaction-time effect on a single outcome measure and a mood-improvement result that didn’t quite reach statistical significance. Thin findings got stretched to cover the entire emotional and cognitive experience of menopause.
The strong evidence for one claim boldly lent to the others.
Who benefits? The powder is cheap, so the money isn’t all inside the tub.
It’s in the authority—the coach, the brand, the social media reel that turns a muscle supplement into an answer for menopausal symptoms. Here, a scoop arrived where the foundational work should have: strength training the muscles, eating enough protein to build them, sleeping enough to recover. These are things creatine supports but cannot replace.
How to hold it: Creatine is real for what it’s proven to do, oversold for the rest. Both at the same time. It’s worth considering for the reason the research supports, not the reasons the algorithm added. And the plain, cheap monohydrate on the bottom shelf is the one the studies have actually used. The premium “women’s formula” in the pink or lavender tin is charging for the marketing, not the molecule. First and foremost, don’t skip the foundational work. Second, carefully consider how and where you spend your dollars.
3. The chatbot is the listener that no fifteen-minute appointment leaves room for.
She types the symptom, the timeline, the thing that’s been waved off in every fifteen-minute appointment—and this time something answers in full paragraphs, unhurried, taking her seriously.
The relief of being heard is immediate. Here, that relief is also the problem.
Whose evidence is shaping this? A June 2026 study found that more than nine in ten adolescents and young adults who asked an AI for mental health advice judged it helpful. The researchers raised concern about what such a high helpfulness rating actually measures, given how chatbots are designed—to agree, to affirm, to keep the conversation going.
A separate BMJ Open analysis pushed five popular chatbots through 250 health questions—many deliberately baited toward misinformation. Out of all 250, the bots declined to answer only twice. The rest of them were answered confidently, footnotes and all. When researchers reviewed those answers, they found nearly half were problematic—misleading at best, harmful at worst. And the footnotes didn’t hold up either. Most of the citations couldn’t be verified.
Helpful-feeling and correct are two different measurements. The tool optimizes for the first.
Who benefits? The AI platform. It earns its keep because the user feels seen, heard, and understood. This brings her back tomorrow.
And the affirmation lands squarely in the old wound—women dismissed for years finally feel met, by something that mostly reflects them back to themselves. A confirmation arrived where a compassionate and curious clinical encounter could have served better. Agreement is not an honest assessment of her own body or anyone else’s she’s worried about.
How to hold it: The warmth that AI is providing, and our draw to it, is a telltale sign of where human interaction is failing. The warmth feels meaningful. But the judgment is unproven. Chats don’t know what they don’t know. They may be useful for shaping a question, but not for settling it. The instrument built to never push back is the last one to trust with an answer you most want to hear and act on. The clinician who challenges or disagrees is doing the part the machine is engineered to skip.
There’s a curious irony here: The machine is holding something only humans can truly give. It’s doing a weak job of something humans have nearly stopped doing altogether—relating to one another—and that anemic job feels like deep, existential relief compared to the emptiness of our human interactions. We end up trusting where trust has not been earned.
That’s our discernment practice this month, applied to three different headlines. Here are three questions to bring to the next claim that arrives sounding certain:
Is this a trial, or amplified testimony?
What, exactly, is this proven for?
Will this source give me a true answer, or just reinforce my beliefs?
There’s one pattern under all three. Something easy arrived where something real belonged—a pill for a conversation, a scoop for the work, a screen for a second opinion. The practice is catching the trade before you make it, so you can see the next claim through a different lens.
If you’re in the Inner Circle, I’ll see you with another Discernment Dispatch at the end of August. If not, I’d love to have you join us.
References:
Dispatch 1
Rhodes JR, Alldredge CT, Elkins GR. Magnitude of placebo response in clinical trials of paroxetine for vasomotor symptoms: a meta-analysis. Frontiers in Psychiatry. 2023 Aug 4;14:1204163. doi: 10.3389/fpsyt.2023.1204163.
The 2023 Nonhormone Therapy Position Statement of The North American Menopause Society. Menopause. 2023 Jun;30(6):573-590. The position statement does not recommend either fexofenadine or famotidine for vasomotor symptoms.
Zhou T, et al. Estimation of placebo effect in randomized placebo-controlled trials for moderate or severe vasomotor symptoms: a meta-analysis. Menopause. 2023 Jan;30(1):5-10. doi: 10.1097/GME.0000000000002094.
Heidelbaugh JJ. Proton pump inhibitors and risk of vitamin and mineral deficiency: evidence and clinical implications. Therapeutic Advances in Drug Safety. 2013;4(3):125-133. doi: 10.1177/2042098613482484. The strongest evidence for absorption interference concerns proton pump inhibitors. For H2 blockers like famotidine, the evidence is more mixed—some studies find associations with B12 deficiency in older adults on chronic therapy, though a 2017 review found the evidence for H2 monotherapy inconsistent.
Dispatch 2
Korovljev D, Ostojic J, Panic J, Ranisavljev M, Todorovic N, Nedeljkovic D, Kuzmanovic J, Vranes M, Stajer V, Ostojic SM. The Effects of 8-Week Creatine Hydrochloride and Creatine Ethyl Ester Supplementation on Cognition, Clinical Outcomes, and Brain Creatine Levels in Perimenopausal and Menopausal Women (CONCRET-MENOPA): A Randomized Controlled Trial. Journal of the American Nutrition Association. 2026 Mar-Apr;45(3):199-210. doi: 10.1080/27697061.2025.2551184. Epub 2025 Aug 25.
Smith-Ryan AE et al. Creatine in women’s health: bridging the gap from menstruation through pregnancy to menopause. Journal of the International Society of Sports Nutrition. 2025;22(1):2502094. Notes explicitly that perimenopausal data remains limited. Lead author has industry ties with creatine manufacturers.
Xu C, Bi S, Zhang W, Luo L. The effects of creatine supplementation on cognitive function in adults: a systematic review and meta-analysis. Frontiers in Nutrition. 2024 Jul 12;11:1424972. doi: 10.3389/fnut.2024.1424972. The meta-analysis pooled 16 RCTs and reported some beneficial effects on memory, attention, and information processing speed in adults. The methodology has been subsequently questioned by the European Food Safety Authority (2024) and in a published Frontiers commentary (2026) for pooling non-independent cognitive test results, which the EFSA concluded inflated the findings. The Xu meta-analysis is still the most-cited recent meta-analysis on creatine and cognition in adults.
Dispatch 3
McBain RK, Cantor JH, Breslau J, et al. AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. 2026 Jun 1. doi: 10.1001/jamapediatrics.2026.2015.
Tiller NB, Marcon AR, Zenone M, Kidd KE, Jeukendrup AE, Master Z, Caulfield T. Generative artificial intelligence-driven chatbots and medical misinformation: an accuracy, referencing and readability audit. BMJ Open. 2026 Apr 14;16(4):e112695. doi: 10.1136/bmjopen-2025-112695.
Cheng M, Lee C, Khadpe P, Yu S, Han D, Jurafsky D. Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence. Science. March 2026. doi:10.1126/science.aec8352 The study tested 11 leading AI models (including ChatGPT, Claude, Gemini, and DeepSeek) across three datasets of social prompts—including thousands from a Reddit community and thousands of statements describing deceitful or illegal conduct. The models affirmed users 49 percent more often than humans did on average. Users trusted the complimentary responses more, preferred them, and said they would be more likely to use the chatbot again.
P.S. One more thing before you go. A reader commented on last week’s piece—asking whether any of this perspective holds when you’re living with serious chronic illness and “just do the basics” plainly isn’t enough. It’s the best question I’ve been asked in a while, and my answer took some time to articulate. Then she came back a day later to say she’d run the two questions on a very well-marketed liver flush protocol and talked herself out of it. That whole exchange is in the comments, and honestly it’s the best thing attached to that piece on the wellness marketplace. Read the thread here →
Again, a beautifully articulated thought provoking piece. I feel compelled to share this-the creatine craze is real. I too, as a woman in the season of life that asks me to maintain my muscle mass, discovered that a prescribed medication that I take, which has afforded me a better quality of life, causes creatine levels to increase. For me, the supplement would be consequential. The lesson here is to be aware of the full ramifications of ANY supplement-despite the promises. We have a responsibility to be educated consumers, albeit easy to follow the Pied Piper.