Side Effects Bryan Johnson Publicly Discussed (and What They Match in the Clinical Literature)

Published May 25, 2026Last reviewed Aug 3, 2026

Image: HealthRX.com AI-generated clinical image

Quick answer. Bryan Johnson has published a changing self-experimentation protocol, but only some adverse effects can be attributed to him from a primary source. The clearest example is rapamycin: Johnson wrote that he stopped it on September 28, 2024 after intermittent skin or soft-tissue infections, lipid abnormalities, glucose elevations, and a higher resting heart rate . His current published protocol lists acarbose and NMN, while an older official archive documents transdermal testosterone as a historical intervention. Those records do not prove that every known drug adverse effect happened to Johnson, and they do not confirm the old page's claims that he personally reported acarbose GI symptoms or currently uses several unapproved peptides. The clinical evidence below separates direct self-report, current or historical use, labeled risk, and trial findings.

At a glance

Directly reported adverse effects / rapamycin-associated infections, lipid abnormalities, glucose elevations, and increased resting heart rate
Rapamycin status / Johnson says he stopped on September 28, 2024 after nearly five years of experimentation
Current published entries / the 2026 Blueprint protocol lists acarbose; the 2026 morning routine lists NMN
Historical entry / the official protocol archive documents transdermal testosterone during caloric restriction
Not verified as current / testosterone, BPC-157, injectable GHK-Cu, and hexarelin were not found in the current official protocol reviewed
Evidence boundary / approved-drug labels and trials describe population risks, not proof of an adverse event in one public figure
Longevity boundary / mouse lifespan findings do not establish human lifespan extension
Attribution rule / distinguish what Johnson reported from what a label warns about and what a trial measured

What Bryan Johnson actually reported

Johnson's most specific first-person adverse-effect account appears in his official January 2025 post, I stopped taking rapamycin. He wrote that he stopped on September 28 after almost five years of experiments that included several weekly and biweekly schedules.

He identified four problems that he and his team suspected were related to rapamycin:

intermittent skin or soft-tissue infections;
lipid abnormalities;
glucose elevations; and
increased resting heart rate.

Johnson also wrote that dose adjustments did not resolve the pattern. That is a sourceable account of his decision and observations. It is still an n-of-one attribution: the post says his team suspected rapamycin after considering other causes, not that a controlled rechallenge proved causality.

The old page instead said he stopped because of unspecified immune-marker changes and biological-age readings, then attached only a social-media profile. His post does discuss an aging-clock preprint, but it separately names the adverse effects that drove the decision. The repaired account keeps those reasons distinct.

Rapamycin: where his report matches established risk

Sirolimus is approved as an immunosuppressant in transplant medicine, not as a drug to extend healthy human lifespan. The current DailyMed sirolimus label carries a boxed warning about increased susceptibility to infection and possible lymphoma or other malignancy from immunosuppression. The label also describes hyperlipidemia, edema, hypertension, abdominal symptoms, anemia, thrombocytopenia, and other adverse reactions in transplant populations.

Those label data make Johnson's infection and lipid observations biologically and clinically plausible. They do not provide an incidence estimate for intermittent longevity dosing. Transplant recipients, background immunosuppressants, target concentrations, and treatment objectives differ substantially from those of otherwise healthy users.

A small randomized pilot assigned 25 generally healthy adults aged 70 to 95 to 1 mg daily rapamycin or placebo for at least eight weeks. Reported possible treatment effects in the rapamycin group included rash, stomatitis, and gastrointestinal symptoms, and several red-cell measures decreased (PMID 29408453). The study was designed for feasibility and safety, not lifespan.

The later PEARL trial tested placebo, 5 mg weekly, or 10 mg weekly compounded rapamycin for 48 weeks in adults undergoing normative aging. Adverse and serious-adverse-event rates were similar across groups, and the primary visceral-fat outcome did not improve. Some prespecified or subgroup health measures changed, but the authors called for broader efficacy work (PMID 40188830).

Another often-cited trial used an mTOR inhibitor related to rapamycin, not rapamycin itself. In older adults, short courses of everolimus improved influenza-vaccine response, while mouth ulcers and other events varied by dose (PMID 25540326). Calling that direct proof of chronic rapamycin safety would collapse important differences in molecule, schedule, endpoint, and duration.

The practical conclusion is narrower than the old page's claim that no randomized healthy-adult trial existed. Human trials now exist, including a 48-week study. They remain too small and short to establish that intermittent rapamycin extends human lifespan or is safe over decades.

Acarbose: verified use, but not a verified personal side-effect report

Johnson's 2026 Blueprint protocol lists prescription acarbose. The page contains inconsistent dose listings in different sections, which is one reason this review does not turn the public protocol into dosing instructions.

The old HealthRX page said Johnson publicly described flatulence and bloating from acarbose. No primary Johnson source supporting that personal adverse-effect attribution was identified in the official materials reviewed. The claim has been removed.

The drug's general GI profile is nevertheless well established. Acarbose inhibits intestinal alpha-glucosidases, delaying carbohydrate digestion. The current DailyMed acarbose label reports abdominal pain in 19%, diarrhea in 31%, and flatulence in 74% of treated participants in U.S. placebo-controlled trials, compared with 9%, 12%, and 29% on placebo. Those trials used 50 to 300 mg three times daily in people treated for type 2 diabetes. The rates cannot simply be assigned to Johnson or to a different schedule.

A Cochrane review found that alpha-glucosidase inhibitors improve glycemic measures in type 2 diabetes but commonly cause gastrointestinal effects (PMID 15846673). The label also notes that abdominal pain and diarrhea tended to return toward pretreatment levels and flatulence tended to lessen during a one-year study. That directly contradicts the old page's statement that GI effects are pharmacologically inevitable at every active dose and never disappear.

Why is acarbose discussed in longevity circles? In the National Institute on Aging Interventions Testing Program, acarbose increased median lifespan by 22% in male genetically heterogeneous mice and 5% in females (PMID 24245565). This is animal evidence. It is not a randomized human longevity result, and it does not convert an approved diabetes drug into a proven anti-aging therapy.

NMN: current use, short human trials, no confirmed Johnson adverse effect

Johnson's official 2026 morning routine lists 500 mg NMN. His official protocol archive also documents an earlier 90-day comparison of NMN with nicotinamide riboside and reports his own intracellular NAD measurement.

Those records establish self-reported use. They do not show that Johnson experienced a specific adverse effect from NMN. The old page was right to avoid inventing one, but its evidence links were badly corrupted: the purported 2022 NMN trial resolved to a silkworm-virus paper, and the purported 2024 review had no valid PMID.

A correctly matched randomized trial assigned 30 healthy adults to 250 mg NMN or placebo daily for 12 weeks. It found an increase in whole-blood NAD-related measures and no obvious adverse effects or abnormal physiological and laboratory findings (PMID 35479740). A separate randomized study gave 1,250 mg daily for four weeks to 31 healthy adults and reported no severe adverse events (PMID 36002548).

A 2023 review of human NMN trials emphasized that the studies were generally small and short and that anti-aging efficacy in humans remains unestablished (PMID 37619764). Short-term tolerability is not the same as proof of multi-year safety, disease prevention, or longer life.

The previous page also used an unrelated obesity-epigenetics PMID to raise a specific occult-cancer-growth theory. NAD metabolism is relevant to cancer biology, but the cited record did not support the claim, and human NMN trials have not established that oral NMN promotes occult cancer. Presenting a mechanistic possibility as a demonstrated personal risk would be misleading, so that passage has been removed rather than relinked to a speculative review.

Testosterone: documented historically, not verified as current

The official Blueprint archive states that Johnson used transdermal testosterone to counter lower testosterone during caloric restriction and targeted a laboratory range. That is historical first-person documentation. The current 2026 protocol reviewed for this page does not list testosterone in its prescription stack.

The old page described Johnson's testosterone use as current and said his published bloodwork proved ongoing hematocrit, PSA, and lipid monitoring alongside therapy. It then linked the cardiovascular section to a parasite-detection paper rather than TRAVERSE. Current-versus-historical status and clinical-trial evidence were therefore both misrepresented.

For the medical context, the Endocrine Society guideline requires symptoms plus consistently low testosterone before diagnosing hypogonadism and recommends monitoring according to patient risks and formulation (PMID 29562364). TRAVERSE randomized 5,246 men with symptoms, repeatedly low testosterone, and cardiovascular disease or elevated risk. Testosterone gel was noninferior to placebo for major cardiovascular events, although atrial fibrillation, acute kidney injury, and pulmonary embolism occurred more often in the testosterone group (PMID 37326322).

Those findings concern indicated replacement in a screened population. They do not prove a longevity benefit, validate a public figure's target range, or confirm Johnson's current treatment status.

Peptides: the old attribution was not sourceable

The previous page said Johnson's published protocol included BPC-157, GHK-Cu, and hexarelin. The current official protocol and the official archive reviewed for this repair did not substantiate that list. The BPC-157 citation was not a wound-healing review at all; it resolved to a case report about ectopic ACTH syndrome in prostate cancer. The hexarelin citation resolved to a sandfly-feeding study.

It would be wrong to preserve a celebrity attribution merely because these compounds are discussed in longevity communities. The repaired page therefore does not claim that Johnson currently uses them.

The general regulatory evidence is still useful. The FDA's safety-risk list for certain compounding substances says it has limited safety information for BPC-157 and identifies immunogenicity, peptide-impurity, and characterization concerns. The same page identifies limited human safety data and immunogenicity concerns for injectable GHK-Cu. These are product-level cautions, not evidence about Johnson.

Why a changing multi-intervention protocol is hard to interpret

Blueprint's current protocol, older archive, product pages, interviews, and first-person updates do not represent one fixed exposure. Entries are added, removed, cycled, or reformulated. The official acarbose page itself shows why dates and versions matter: dose figures differ between sections.

That creates three attribution problems:

A compound appearing in an archive does not prove current use.
A known label adverse effect does not prove Johnson experienced it.
A biomarker changing during a multi-intervention protocol does not identify one cause without a stronger design.

The best use of Johnson's public record is to generate testable questions. The strongest adverse-effect statement on this page is his own dated rapamycin account. The acarbose, NMN, testosterone, and peptide sections deliberately use different language because the underlying evidence is different.

What the source audit corrected

This repair removed one malformed PubMed destination and five valid-looking but unrelated numeric PMIDs. The unrelated records concerned silkworm virus replication, obesity epigenetics, river-water parasite detection, ectopic ACTH syndrome in prostate cancer, and sandfly feeding.

It also corrected or removed these unsupported statements:

that Johnson reported acarbose flatulence and bloating;
that acarbose GI effects are inevitable and persist at every active dose;
that a 2022 NMN trial improved insulin sensitivity in overweight adults at the cited PMID;
that a 2024 NMN/NR systematic review existed at the malformed destination;
that a cited cancer paper supported an occult-malignancy warning about NMN;
that testosterone use was confirmed as current in 2026;
that the wrong PMID was the TRAVERSE cardiovascular trial;
that Johnson's current published protocol confirmed BPC-157, GHK-Cu, and hexarelin; and
that a mouse or preclinical finding established a human longevity benefit.

Frequently asked questions

Which side effects did Bryan Johnson directly say he experienced on rapamycin?

In his official January 2025 post, Johnson identified intermittent skin or soft-tissue infections, lipid abnormalities, glucose elevations, and increased resting heart rate. He said he stopped rapamycin on September 28, 2024 after dose changes did not resolve the pattern.

Did Bryan Johnson report acarbose GI side effects?

No primary official source reviewed for this page showed Johnson personally reporting flatulence or bloating from acarbose. His current published protocol lists the drug, and GI effects are common in the approved diabetes trials, but those are separate facts.

Does Bryan Johnson currently take NMN?

His April 2026 morning-routine post lists 500 mg NMN. Human trials suggest short-term tolerability and increased NAD-related measures, but they do not establish that NMN slows aging or extends human lifespan.

Does Bryan Johnson currently use testosterone?

The official archive documents historical transdermal testosterone use during caloric restriction. The current 2026 prescription stack reviewed here does not list it, so this page does not describe it as current therapy.

Does Blueprint prove that rapamycin or acarbose extends human life?

No. Rapamycin and acarbose have substantial animal-longevity literature, and small human rapamycin safety studies now exist. No randomized trial has shown that either drug extends human lifespan in healthy people.

Can one drug be blamed for a biomarker change in a large protocol?

Not from timing alone. A dated stop-and-rechallenge pattern can strengthen an inference, but concurrent drugs, supplements, diet, training, illness, measurement variation, and protocol changes can all affect the result.