Mallory's Notebook — June 2026
Short notes from the practice: what I'm seeing in the exam room, what the new data changed, and what I'm paying more attention to this month.
- 01Compounded semaglutide quality is genuinely improving — but only at pharmacies investing in API verification.
- 02Perimenopause is now my second-most common new-patient reason. Most have been told their labs are normal.
- 03The protein-and-training case for GLP-1 patients is no longer controversial. The studies caught up.
- 04I'm seeing more late-diagnosed adult ADHD in women, often unmasked by perimenopause.
A few notes from May.
On compounded GLP-1s
The quality gap between reputable and disreputable compounding pharmacies has widened in the last six months, not narrowed. The good pharmacies are publishing certificates of analysis, confirming base semaglutide rather than salts, and operating well within 503A rules. The bad ones are still doing flat-fee subscriptions on plain compounded product with no individualization, which the FDA has been clear about.
I changed one of my pharmacy partners in May for exactly this reason. If you're a current patient, nothing about your prescription changed except the label.
On perimenopause
Perimenopause is now the second-most common reason patients book a new-patient visit, after weight management. Almost all of them have been told their labs are normal, often multiple times.
The clinical pattern is consistent enough now that I am comfortable saying: if a woman between 38 and 52 has new sleep fragmentation, new anxiety, or new joint pain, perimenopause belongs on the differential before "stress" does. A normal TSH and a normal estradiol on a single draw do not rule it out.
I published a longer piece on this on May 7.
On GLP-1s and muscle
The 2025 data on resistance training during GLP-1 therapy is now consistent enough that I think the conversation is settled. Protein targeting (0.8–1 g/lb goal body weight) plus 2–3 weekly resistance sessions preserves lean mass at rates similar to non-pharmacologic weight loss.
The patients I worry about are the ones losing weight on telehealth platforms with no protein guidance and no body composition tracking. That is where the muscle loss stories come from, and it is preventable. Full piece here.
On adult ADHD
Late-diagnosed ADHD in women — often unmasked or worsened by perimenopause — is now a regular pattern. The intersection matters: treating one without considering the other usually fails. Stimulants on top of unaddressed estrogen decline tend to disappoint. Hormones without addressing real attentional dysfunction tend to disappoint, too.
I'm getting better at sequencing this. Usually I start with the more reversible piece (sleep, iron, hormones) and re-evaluate ADHD symptoms once those are stable. Sometimes the answer is both. Sometimes it's just one.
On reading
Two papers I'm still thinking about from May: the SELECT trial follow-up on semaglutide cardiovascular outcomes at four years, and a smaller but interesting paper on testosterone and verbal memory in postmenopausal women. Happy to talk about either in your next visit if you're curious.
— M.
Ready for the conversation?
Have a real visit with Mallory.
Telehealth visits with one clinician, continuity between visits, and no rotating staff.
