I personally would prioritize a statin, ezetimibe, glp-1, sglt-2 inhibitor, possibly low dose doxycycline, making sure vitamin B12/Omega 3’s/Vitamin D are optimized, tadalafil, low dose lithium as being a pretty good stack. Some of these items will prove to useless, but some of them will likely pan out. I just can’t say which as of today.
Lifestyle is critical in addition, with sleep, exercise, diet, stress all have probably a lot more importance that much of the items above.
The only thing you might be missing is HRT for women — if caught early enough in the transition to perimenopause/menopause. Estrogen is the foundation for everything to women.
Also ChatGPT warns me against doxy even in small subclinical doses, as potentially damaging to mitochondria, which are essentially ancient bacteria who’ve evolved laterally.
I agree with both Male and Female HRT in this area - although we have data that looks concerning when female HRT is started late - so there is a potential need to nuance this with trying to start this around menopause.
On the doxycycline, 20 mg BID … I don’t share the concern - but at 100 mg bid long term or even 100 mg daily, this is an interesting area.
Here is what Vera-Health.ai says on this issue:
At sub‑antimicrobial dosing (doxycycline 20 mg BID) taken long term, there is no direct human evidence that it causes clinically meaningful, persistent mitochondrial dysfunction, but mechanistically it can inhibit mitochondrial translation, so a small/individual‑susceptibility risk cannot be excluded.
Yes absolutely right re: timing of HRT. It’s what I meant when I said “if the transition is done early enough.” I’ve had a very consultative, Socratic style discussion with ChatGPT on my health interventions and have found it orders of magnitude more useful than the way I’d been used to employing it before — i.e., hit and run digests in single questions / topics, as I often see done here too. It is usually very conservative about suggesting interventions for me and seems more “comfortable” being reactive in judging what I myself come up with, but with respect to HRT it’s almost insisting that I go on it asap. Funny thing is I haven’t felt any symptoms so far (I’m 39) but if thinks that because I’ve nursed 5 babies for a long time and am still nursing my toddler, I am de facto hypoesterogenic and have had a bit too much area under the curve over a lifetime so far in that condition so especially given my ApoE3/4 genotype, the single biggest lever I can pull to protect my brain is get on that HRT asap.
I’ve been amazed at how nuanced the recommendations have been though — it has consultatively come up with a stack for me that’s highly personalized on what it knows about my physiology and psychology and it’s been working great this past month since I started (best sleep in years, measured, much stronger nails, great skin, etc.) but I wonder how much of my stack is transferable to anyone else. It honed in on my oura data and 23&me data too. For me for example it is actively discouraging PDE-5 inhibitors saying they’re too blunt for me now, but perhaps later down the road, but it is green lighting l-citrulline. It’s got me taking on a @John_Hemming lite protocol with citrate salts indefinitely for kidney protection (I am prone to uric acid stones and have the low klotho genotype). I’ve really thrown the kitchen sink at it and got rid of all my supplement / pharma fomo, streamlining everything into sure, take it / maybe later / never touch buckets.
Re: doxy, it would tell me (and practically has when I brought up that reasoning) that absence of evidence of harm is no evidence of absence of harm. Its mechanistic reasoning is very persuasive.
Saw the attached article and asked Perplexity to review it and answer a few questions.
As one of my interests is how senescent cells affect our overall health and health span, I had to ask about that as well. Especially since the author and “patient” Dr. Brent Beasley, notes that a Covid infection was a major trigger but “not the cause”.
I believe that senescence is a significant cause of all types of dementias.
If you are not killing your zombie cells, they will get you.
slowing_alzheimers_senescence-factor pdf.pdf (469.9 KB)
This seems to be the increasing realization… infections (and the associated inflammation) are much worse for the brain than we thought: Viral Pneumonia Accelerates Brain Aging
Indeed - on my video about immunizations - the big benefit is keeping someone from having a major inflammatory response as that injures all organs, but it seems like the brain is quite susceptible.
So I do recommend pretty much all the vaccines. I suspect with Covid, if you take Paxlovid early - probably just a good.
My experience thus far with patients on Rapamycin has been pretty favorable in regard to not getting horribly unwell with viral illnesses. Considering doing a similar protocol to the TRIIM trial to improve T cell immunity also should be on the plate as folks get older.
On @Steve_Combi 's article, yes, I have patients positive for Beta Amyloid 42:40 ratio and p-Tau 217 with no symptoms or very vague MCI. It is something to track to see if interventions are working and improving these test results, or slowing predicted worsening in these values over time. This is where the DORA agents are interesting as they have been linked with decrease in the isoforms of Beta Amyloid - but not necessarily the ratio we track.
Dr. Fraser, Would you please share your opinion of Paraspinal Etanercept. There are some studies showing that it can reverse MCI/early AD. It is approved for rheumatoid arthritis but is starting to be used off label (not approved by the FDA) for AD and seems to show promise. Do you think this is worth chasing, and if so, how/where to find a practitioner who would prescribe and inject (into the spinal column). Would appreciate your view –
I think I’ll go with Vera-Health.ai’s review of this … and apart from the expense, you’d have a very hard time finding someone to inject it; and the rationale for paraspinal versus simply subcutaneous makes little sense. This drug gets widely distributed through the body (apparently not crossing the BBB or into the spinal column) slowly, but through lymphatic uptake, which would have it be moving away from the spine anyway.
I think if one were compelled to do this, it would be hard to find a physician to prescribe it for this indication given it has more significant risks that anything that I would prescribe that I’d lump into “speculative therapy.” One of my rules for this, is that it has to be very low risk for adverse outcomes.
Clinical evidence for perispinal/paraspinal etanercept in Alzheimer’s disease is limited to uncontrolled human reports and is not supported by randomized controlled trial evidence; it should not be used outside a clinical trial.
Clinical Guideline
Do not offer etanercept (including perispinal/paraspinal administration) to treat Alzheimer’s disease outside research settings because there is no high-quality evidence of clinical benefit and biologic therapy carries meaningful infection risk. 1 I
Thank you so much for your reply.
I guess for now, in addition to “behavior” will stick to lithium, tadalafil, intranasal insulin, and a few other supps like bacopa monnieri --but will continue to explore. MCT oil???
At the recent UC Berkeley longevity conference I spoke to one of the local scientist attendees who mentioned his wife is APOE3/4 and was concerned about her risk, so took a proteomics test provided by the following company: called “Molecular You”. https://www.molecularyou.com/
Here is a link to their website product page. Advanced Biomarker Insights | Biomarker Testing Service Company where they describe their services looking at 190+ biomarkers.
It looks like they charge $1,100 list price for the one time test (and maybe annual subscriptions), which includes a home visit by a phlebotomist to draw blood.
Note these guys DO NOT test for genetic markers like APOE4, but only proteomic ones which would be activated, in theory, as a result of, say, APOE4—but not necessarily if there are other mitigating factors.
The result of the proteomics evaluation was that, despite the E3/4 status, they concluded she was extremely low risk of alzhiemers due to all the positive measures evident from the proteomics test (I think she was in excellent health, regular exerciser, good blood markers, etc. too).
So, perhaps another way to look at your E4 risk profile with greater information.
Anything new on if Rapamycin are still on for many of those SV tech people? If not on, what they opted to? Metformin? GLP1?
More good news on rapamycin:
Carriers of the apolipoprotein E4 (APOE4) allele often develop cerebrovascular dysfunction and broader systemic alterations decades before the onset of Alzheimer’s disease (AD) pathology or clinical symptoms. Early interventions that can improve these functions may help delay or slow AD progression. In this study, we repurposed rapamycin (sirolimus), an FDA-approved medication with anti-aging properties, to target APOE4 associated multisystem dysfunction. We conducted a single arm trial of low dose rapamycin (1 mg/day for 4 weeks) in cognitively normal adults aged 45 to 65 years, prior to detectable AD pathology, stratified by APOE genotype. The primary outcome was cerebral blood flow (CBF), while secondary outcomes included plasma metabolomics, inflammatory cytokines, AD biomarkers, and gut microbiome composition. Twenty-three participants completed the study, including nine APOE4 carriers and fourteen non-carriers. Rapamycin significantly increased CBF in APOE4 carriers, with increases exceeding 15% across multiple brain regions, and improved metabolic and inflammatory profiles with minimal adverse effects. In contrast, non-carriers showed no significant change in CBF and exhibited distinct physiological responses, highlighting genotype dependent effects. These findings suggest that rapamycin may mitigate early cerebrovascular and systemic dysfunction in APOE4 carriers and support a precision medicine approach in which therapeutic response is influenced by genotype.
It is interesting the items that can increase CBF, and the interventions seem more effective in those with an ApoE4.
Aerobic exercise, Strength training, Omega 3 fatty acids, Anserine/Carnosine, Ketogenic diet plus HIIT, Sleep optimization, Nitric oxide optmization, NAD precursors, and for those who have enough blood pressure - ARBs or ACE inhibitors that get into the brain (telmisartan is my favorite), Mediterranean diet … among others. The discussion on hypercapnia and deep breathing also comes into play.
I suspect most of us are doing a combination of many of these.
Bisphosphonates and the risk of dementia in patients with osteoporosis or fragility fracture: A population-based study in Hong Kong
https://alz-journals.onlinelibrary.wiley.com/doi/10.1002/alz.70503
Pop-sci article:
Common Osteoporosis Drugs Linked to Lower Alzheimer’s Risk in 120,000-Person Study
“Researchers at the University of Hong Kong analyzed electronic medical records from more than 120,000 people age 60 and older. NBP users had a 16% lower risk of developing Alzheimer’s disease and related dementias than untreated patients and a 24% lower risk than people who received other osteoporosis medications.”
This is a fascinating area, given all the controversy of so many of my patients who have no interest in taking things like Fosamax (alendronate) due to all the potential horrible side effects (which are actually significantly overstated).
My first time I appreciated this class as a “longevity medication” outside of simply osteoporosis treatment was this discussion in the European Society of Medicine in 2024, of Geroscience Drugs.
I’m in agreement with the SGLT2’s, GLP’s, Rapamycin … as I go down the rest of their table, I’m not so sure … but the 3rd item, Bisphosphonates … like Fosamax … that was an item I didn’t expect to see. This article adds to my interest in this class.
UpdatedLongevityDrugListPaperFeb2024.pdf (278.0 KB)
Yes, there are threads on here where bisphosphonates as geromodulators were discussed. One of mechanisms that was speculated about was impact on the CV system. So ultimately ACM is lower for folks on these drugs. And yes, they are effective against osteoporosis. You are also right that people are very fearful of these drugs based on exaggerated risk assessments. However, while I am usually pretty immune to such panics, I admittedly find one aspect of these drugs rather disquieting: dental issues. It’s not really about the jaw necrosis - which is rare. Rather something else: apparently, once you’ve been on a course of bisphosphonates, the bones around your teeth are permanently altered, and not for the better. You are stuck for life - and if you anticipate dental issues, which is common as you age, your dental bone health will be compromised going forward, and as an elderly person you have very little room for error when it comes to teeth. That stopped me dead in my tracks - especially that there are probably some dental implants still in my future. So, while I am intrigued by these drugs, I am electing to pass on them for the time being. YMMV.
Well, that’s concerning.
I’m an elderly person who’s been on bisphosphates on and off since 2017. I have no original teeth. Rather, crowns, bridges, and implants. Had lots of dental work done over that period. So far, at least, no indication that bispho’s had any negative effect. N of 1 of course. YMMV.
I asked gpt to chase down the studies that support this assertion. Results follow. If the studies behind this claim aren’t considered in the ai search below, please share.
As always, if you’re unlucky enough to be the ‘1 in 2000’ to experience a rare side (severe) effect, you don’t get a do-over. So a good reason to be cautious.
The claim in that post is considerably stronger than the evidence supports. I can find solid evidence that bisphosphonates can produce long-lasting suppression of bone remodeling in alveolar bone, and that drug remains associated with the skeleton for years. I cannot find evidence showing that an osteoporosis-dose course causes the bone around the teeth to become permanently damaged or permanently inferior for dental work.
The distinction matters: persistent pharmacologic effect ≠ permanent structural damage.
Where the claim probably comes from
Evidence What it actually shows Very long skeletal retention Alendronate binds hydroxyapatite and has an estimated terminal skeletal half-life of >10 years. But that figure is a pharmacokinetic estimate; a later pharmacology review cautions that the often-quoted 10-year number is extrapolated and shouldn’t be interpreted too literally. Stored drug is largely inactive until the bone surface is resorbed. (dailymed.nlm.nih.gov) Alveolar remodeling suppression In beagles given clinically relevant alendronate for 3 years, alveolar mandibular turnover was about 75% lower, and focal microscopic necrotic areas occurred in 25–33% of treated animals versus none of the controls. Importantly, none developed clinically exposed jaw bone. This is probably one of the strongest experiments supporting concern specifically about alveolar bone. (pmc.ncbi.nlm.nih.gov) Abnormal human jaw tissue in MRONJ Biopsies from people who had actually developed bisphosphonate-related ONJ show abnormal remodeling, thicker trabeculae, fewer Haversian canals, osteocyte loss, etc. But these are selected diseased patients, often receiving high oncology doses, so they cannot establish that ordinary bisphosphonate users have the same pathology. (pubmed.ncbi.nlm.nih.gov) Orthodontic effects Human case reports and small cohorts show slower tooth movement and sometimes sclerotic alveolar changes. A 2026 umbrella review says the effect on orthodontic tooth movement is consistent, but human evidence remains sparse and weak. (link.springer.com) The beagle study is particularly interesting because the alveolar portion of the jaw normally remodels much faster than non-alveolar mandibular bone. That provides a plausible mechanism for why the jaw is unusually vulnerable to potent antiresorptive therapy. But it still does not establish irreversible damage. (pmc.ncbi.nlm.nih.gov)
Human evidence is much more reassuring
There are several studies that actively cut against the “your dental bone is compromised forever” interpretation.
A randomized trial of 335 people with periodontal disease treated for two years with weekly alendronate found no excess alveolar bone loss or ONJ; among subjects with low mandibular BMD, alendronate actually reduced alveolar bone loss. (pubmed.ncbi.nlm.nih.gov)
A radiographic study of women treated with bisphosphonates for an average of roughly 4–5 years found no significant alteration of alveolar or basal jaw trabecular structure compared with untreated women. (pubmed.ncbi.nlm.nih.gov)
More directly relevant to dental surgery, a prospective study of women with osteoporosis taking alendronate or yearly zoledronic acid found that tooth-extraction sockets and overlying mucosa healed successfully; by 90 days all patients had mucosal healing, without requiring discontinuation of bisphosphonate treatment. (pubmed.ncbi.nlm.nih.gov) Another prospective study of 132 oral-bisphosphonate users did find slower socket healing after >5 years of treatment, but no cases of ONJ occurred. (pubmed.ncbi.nlm.nih.gov)
And the major randomized zoledronic-acid osteoporosis trial is especially useful here: 7,714 women, zoledronic acid 5 mg yearly versus placebo for three years. There was one potential ONJ case in the zoledronic-acid group and one in the placebo group; both resolved. (pubmed.ncbi.nlm.nih.gov)
What about future implants?
This is probably the most important implication of the post.
The 2025 International ONJ Taskforce systematically reviewed implant outcomes in osteoporosis patients receiving antiresorptives. Their conclusion was that current evidence does not show increased dental-implant failure, and they suggest that antiresorptive treatment does not have to be stopped before an implant. They explicitly say implants can generally be placed in osteoporosis patients receiving bisphosphonates, although the evidence quality for implant outcomes is low. (pubmed.ncbi.nlm.nih.gov)
There is nevertheless a small MRONJ signal. The accompanying 2025 meta-analysis found a pooled post-implant MRONJ rate of about 0.5% among antiresorptive-treated osteoporosis patients; one adjusted cohort suggested bisphosphonate exposure added about 3 MRONJ cases per 1,000 patients. Implant failure itself was not increased in the primary patient-level analysis. (pubmed.ncbi.nlm.nih.gov)
Another 2025 meta-analysis illustrates how uncertain the evidence remains: it found an implant-level failure RR of 1.74, but no patient-level increase (RR 1.01), no significant increase in marginal bone loss, and rated the evidence very low certainty. (pubmed.ncbi.nlm.nih.gov)
The “permanent” part does not hold up
There is excellent evidence that effects persist after treatment stops. In FLEX, after five years of alendronate followed by five years off treatment, bone-turnover markers gradually rose toward pretreatment levels. (pubmed.ncbi.nlm.nih.gov) With zoledronic acid, after three annual doses, measurable antiresorptive effects persisted for another three years, but BMD began declining and PINP rose—again demonstrating an offset of effect, rather than permanent suppression. (pmc.ncbi.nlm.nih.gov)
So I would rewrite the post’s assertion this way:
Bisphosphonates become incorporated into bone and can influence remodeling for years after treatment. Because alveolar bone has unusually high turnover, dental extraction and implant surgery carry a small persistent MRONJ risk, particularly with prolonged treatment and high oncology doses. There is no evidence that osteoporosis-dose bisphosphonates permanently render alveolar bone unhealthy or permanently compromise a patient’s ability to receive dental implants.
That’s substantially different from “you are stuck for life.”
For osteoporosis-dose zoledronic acid specifically, the evidence is substantially more reassuring than the post implies. The very alarming jaw literature is disproportionately derived from cancer patients receiving much larger and more frequent IV doses. AAOMS currently estimates MRONJ risk from bisphosphonates used for osteoporosis at roughly 0.02–0.04% overall, while still acknowledging that invasive dental procedures and longer exposure increase it. (aaoms.org)
Confidence: high that the “permanently altered and compromised for life” statement is not supported by the available human evidence. The more defensible claim is long-lived alteration of remodeling with a small, nonzero dental-surgical risk.
If useful, I can next dig specifically into zoledronic acid/Reclast and dental implants after 1, 2, or 3 annual 5-mg infusions; that would narrow this considerably to the exposure pattern most relevant to osteoporosis treatment.
This paper references earlier reports of terminal life of 11 years. This means complete elimination at more than 20 years. If you are starting bis therapy in your 70’s, that’s essentially lifelong persistance, if we assume a 90+ year lifespan. Of course one may claim it’s excessive extrapolation and furthermore perhaps it doesn’t matter for dental work like implants, or as suggested the downside is statistically tiny, but it’s still not a positive in my book. I’d rather not have it in the bones around my teeth for possibly 20+ years. Like I admitted, it’s probably excessive concern, but I am squicked out by bone alterations that last multiple decades in my late years, as to me it spells “lifelong”.
Pharmacokinetic considerations in determining the terminal elimination half-lives of bisphosphonates
https://pubmed.ncbi.nlm.nih.gov/17523760/
“Results: Data were sufficient for analysis of pharmacokinetics in 11 patients. A mean t((1/2)gamma) of approximately 11 years was reported previously, based on analysis of data between days 240 and 540.”
Yes, you can lose teeth even with the best of care. The best and most expensive option is a dental implant. My dental surgeon’s list of no-no drugs includes bisphosphonates.
