Hormone Therapy in Aging: Replacement vs Optimization

I had three patients this week who all wanted hormones, and I genuinely wasn't sure where the line is between reasonable replacement and selling someone something.

That line is actually the whole subject. Replacing a documented deficiency back into the physiological range is ordinary endocrinology. Pushing a hormone above that range in someone who isn't deficient is a different intervention — different risk profile, usually no outcome evidence.

So walk me through the axes. Where is the evidence strongest?

Menopausal hormone therapy — estradiol and progesterone. Multiple randomized trials, position statements from NAMS and the Menopause Society, approved products. The Women's Health Initiative scared a generation of clinicians, but the relevant finding was timing: starting oral conjugated equine estrogen a decade after menopause in older women is not the same intervention as treating a symptomatic 51-year-old.

I still have colleagues telling perimenopausal women they're too young for hormones.

Which is exactly wrong. Early perimenopause is actually erratic elevated estradiol with a failing luteal phase — the symptoms come from that instability, not simply from deficiency. Progesterone drops first. The indication for treatment is symptoms plus appropriate timing, not a specific FSH number.

What about testosterone in men? I had a patient who already had a year of injectables prescribed without an LH measurement.

No LH means no diagnosis. Hypogonadism requires low testosterone on two morning fasting samples plus symptoms, and you must exclude secondary causes — elevated LH points to primary gonadal failure, low or normal LH in someone with low testosterone means something central is driving it and could be reversible. The Testosterone Trials showed modest benefit for sexual function and bone density in older hypogonadal men, but they were not powered for cardiovascular endpoints.

And the cardiovascular signal — how worried should I be?

TRAVERSE was a randomized trial designed to answer that; it didn't show increased major adverse cardiac events in men with hypogonadism and high cardiovascular risk. But the population matters — that's replacement to physiological range, not supraphysiological optimization, and the men had confirmed hypogonadism.

My third patient wanted growth hormone because his IGF-1 — that's insulin-like growth factor, a downstream marker of GH secretion — was in the bottom third for age.

Bottom third for age is not a deficiency. Adult GH deficiency is a specific pituitary diagnosis with a formal stimulation test. More importantly, the biology of longevity actually points the wrong direction here — reduced GH signaling extends lifespan in multiple animal models, and GH receptor deficiency in humans is associated with reduced cancer and diabetes incidence. Excess GH signaling is acromegaly. This is mechanistic and epidemiological data, not a trial, but the direction is not favorable.

So what do I actually tell him?

That being in the lower range for a hormone that probably shouldn't be high is not an indication. If he has symptoms that make you suspect pituitary disease, refer. Otherwise, no.

What about DHEA, thyroid in older patients — I feel like those come up constantly.

TSH rises with age normally — a TSH of 4.5 in an 80-year-old is near their age-specific norm, not subclinical hypothyroidism. Randomized trials of treating mild TSH elevation in older adults were null for symptoms. DHEA replacement trials show small biomarker effects, no convincing clinical benefit. And adrenal fatigue is not a recognized diagnosis — the Endocrine Society has said that explicitly.

So the thing I should carry out of this?

Document deficiency before you treat, confirm it's not reversible, replace to physiological range only, and monitor. Any clinic describing supraphysiological dosing as optimization is using replacement language to sell you a different product.