Off-label use — what it is, what it isn't

Off-label prescribing is the legal use of an FDA-approved drug for an indication, dose, or population that is not on the approval label. It is common, accounts for roughly one in five outpatient prescriptions in the U.S., and is not — by itself — a marker of low standards. Many cancer regimens, most pediatric psychiatry, and a large share of chronic-pain prescribing live off-label because the trials that would update the label have not been done.

What off-label is not: an FDA blessing. The label reflects the indications the manufacturer paid to study and the FDA approved. The absence of an indication from the label does not mean the drug does not work for it; it means the trial that would prove it has not been run, or has not been submitted, or has been but did not meet primary endpoints. All three are different epistemic situations.

The wellness conversation increasingly happens in the off-label zone. Metformin for healthy aging. Low-dose naltrexone for chronic inflammation. GLP-1s for body composition in lean adults. Rapamycin for healthspan. These are real drugs with real safety profiles, used for indications where the trial evidence is thinner than the on-label case. The honest move is to say so out loud and assess each on the evidence it actually has.

GLP-1 receptor agonists — the cardiometabolic reset

GLP-1 (glucagon-like peptide-1) receptor agonists — semaglutide, liraglutide, tirzepatide — entered the formulary as type-2 diabetes drugs and have, in five years, become the most consequential class in cardiometabolic medicine since statins.

The pivotal data on weight: STEP-1 showed ~15% weight loss at 68 weeks with weekly semaglutide [Wilding 2021]. SURMOUNT-1 pushed that past 20% with tirzepatide [Jastreboff 2022]. The pivotal data on hard endpoints: SELECT showed a 20% reduction in major adverse cardiovascular events in adults with obesity and pre-existing cardiovascular disease [Lincoff 2023]. FLOW was stopped early for kidney benefit [Perkovic 2024]. These are large effects on outcomes that, historically, have moved only modestly with the best interventions in their category.

The caveats, stated plainly: lean-mass loss tracks total-mass loss in roughly a one-for-three ratio, which is why resistance training and adequate protein (1.2–1.6 g/kg) are not optional adjuncts. Rebound after discontinuation is well-documented; most of the weight returns within a year of stopping. Gastrointestinal side effects are dose-dependent and titration-sensitive. And the access economy — both the on-label prescription channel and the compounded-pharmacy channel — has shifted the social and financial dimensions of treatment in ways the trials did not measure.

GLP-1s are the first class in a generation where the trial outcome and the consumer claim are actually aligned. That alone is worth marking.

Metformin — TAME and the longevity bet

Metformin is a biguanide approved for type-2 diabetes since 1957 in France and 1995 in the U.S. It has one of the longest real-world safety records in modern pharmacology. The longevity case is built on observational data — diabetic patients on metformin appearing to have lower all-cause mortality than non-diabetic controls in some cohorts [Bannister 2014]. The interpretation is contested; selection effects in observational diabetes-cohort data are notoriously hard to untangle.

TAME (Targeting Aging with MEtformin) is the planned trial that would settle the question — a multi-center, placebo-controlled trial of metformin in non-diabetic older adults, with a composite endpoint of age-related disease incidence. As of this writing, TAME has not completed enrollment at full scale; the design is widely cited, funded incompletely, and represents the kind of trial geroscience as a field has struggled to mount. Until it reads out, the case for metformin in healthy adults is mechanistic and observational, not outcome-confirmed [Barzilai 2016].

There is also a counter-signal worth tracking. Metformin appears to blunt the training-induced mitochondrial adaptation in some trials in older adults, including the MASTERS trial [Konopka 2019]. If you are layering metformin onto a serious training program, the interaction is non-trivial and the evidence is, at minimum, ambiguous.

Statins — the LDL conversation

Statins inhibit HMG-CoA (3-hydroxy-3-methylglutaryl-coenzyme A) reductase, the rate-limiting enzyme in hepatic cholesterol synthesis, and reduce circulating low-density lipoprotein (LDL) cholesterol by roughly 30–55% depending on agent and dose. The trial base for secondary prevention — patients with established cardiovascular disease — is one of the most replicated in modern medicine; the Cholesterol Treatment Trialists meta-analyses cover hundreds of thousands of patient-years [CTT 2010].

The conversation gets harder in primary prevention. Number-needed-to-treat to prevent one cardiovascular event in low-risk middle-aged adults runs into the hundreds. Side-effect frequency — particularly myalgia — is real but is also blurred by the nocebo effect in open-label populations, as several blinded crossover trials have shown. The honest reading: in established disease or high ten-year-risk patients, statins are foundational. In low-risk primary prevention, the call is genuinely closer than the guideline-versus-wellness-podcast cross-fire suggests.

I am not a statin-skeptic and I am not a statin-evangelist. I think the LDL hypothesis is real, the trial data in secondary prevention is overwhelming, and the conversation in low-risk primary prevention deserves more nuance than either side currently offers.

SSRIs and the modern psychiatric toolkit

Selective serotonin reuptake inhibitors (SSRIs) are the dominant first-line pharmacotherapy for major depressive disorder and several anxiety disorders. The trial base, taken as a whole, supports a modest but real average effect size relative to placebo in moderate- to-severe depression, with the gap widening at the more severe end of the spectrum [Cipriani 2018]. The effect is real; the effect is also smaller than the cultural narrative around these drugs has often suggested.

Sexual dysfunction, emotional blunting, and discontinuation symptoms are well-documented side effects that the older marketing literature underweighted and the more recent clinical conversation has begun to treat with the seriousness they deserve. The discontinuation syndrome is real, can be protracted, and tapering strategies have become more sophisticated over the past decade.

Adjacent to the SSRI conversation: the renewed clinical research into psychedelics (psilocybin, MDMA-assisted therapy) for treatment-resistant depression and PTSD, where phase-2 and early phase-3 data have shown effect sizes larger than typical antidepressant trials. Regulatory paths are evolving. The evidence is promising; the access model is still being built.

Hormone replacement — testosterone and menopause

The hormone replacement conversation suffered a long hangover from the Women's Health Initiative (WHI) results in the early 2000s, which were widely interpreted as showing harm and have since been reanalyzed extensively. The current consensus, reflected in updated guidelines from the North American Menopause Society and others: in symptomatic women within roughly ten years of menopause onset, the benefit-risk balance for hormone therapy on quality of life, bone, and vasomotor symptoms is favorable, with the absolute risks substantially smaller than the WHI headline implied [NAMS 2022].

Male testosterone replacement therapy (TRT) has its own evidence base. In men with documented biochemical hypogonadism and appropriate symptoms, the trial data supports modest improvements in sexual function, mood, bone density, and lean mass; the TRAVERSE trial provided the first large randomized cardiovascular safety signal at the population level [Lincoff 2023b]. The consumer "low-T clinic" economy has expanded the use case well beyond the trial population, with predictable results.

The compounding pharmacy economy

Compounding pharmacies prepare customized formulations of drugs that are otherwise commercially available, and have become the dominant channel for the off-label and shortage-driven side of the pharmaceutical wellness conversation. Compounded semaglutide, compounded tirzepatide, compounded testosterone, and compounded peptides all flow through this channel.

The regulatory framework around compounding is real but lighter than the FDA approval process for finished drug products. Quality control varies meaningfully across compounders. The economics — particularly during the GLP-1 supply shortage — opened a large gray-market marketplace that the manufacturers, the FDA, and several state medical boards have been actively responding to.

None of this means compounded products are unsafe per se; reputable 503A and 503B facilities exist and the channel has legitimate clinical uses. It does mean the variability — in identity, purity, and potency — is higher than the equivalent on-label product, and the consumer is doing diligence work that the FDA approval pathway would otherwise have done.