How to read a free testosterone trend line.
Total testosterone is the number on the report. Free testosterone is the one that matters. The first is easy to measure and easy to misread. The second tells you almost everything — if you know what you're looking at.
01 — Why free, not total
Total testosterone counts every molecule in your blood, including the 98% bound to SHBG and albumin and therefore unavailable to tissue. Free testosterone counts only the unbound fraction — the part that can actually do something. Two people with identical total numbers can have wildly different free values, and the symptoms track the free number, not the total.
02 — The shape, not the number
One free T value is almost meaningless. Three values, taken at the same time of day, over a 90-day window, is a trend line. The trend line tells you whether the protocol is working. The single value tells you almost nothing.
If you remember nothing else: never make a protocol change off a single data point. Always wait for the trend.
03 — Sample size
The minimum useful sample for free T is three data points spanning at least 60 days. Five is better. Anything less and you're reacting to noise. The within-subject variability of free T on a fixed protocol is roughly ±12% — meaning a single 8% drop is well within noise.
04 — Seasonality
Free T has a real, repeatable seasonal pattern. It peaks in late summer in most adults and troughs in late winter. The amplitude is small (about 8–10% peak-to-trough), but it's enough to make a December reading look “worse” than an August one when nothing has actually changed.
If you're comparing across more than four months, control for season. Compare January to last January, not January to last August.
05 — Confounders
- Sleep debt — a single bad night can drop free T by 20% the next morning. Don't draw labs the day after a red-eye.
- Acute illness — anything that spikes inflammation suppresses gonadal output. Wait two weeks past any cold or flu.
- SHBG drift — if SHBG moves significantly, the free-from-total calculation is unreliable. Insist on a direct equilibrium dialysis measurement, not a calculated free.
- Time of day — peak is between 7am and 9am. Always draw in that window, always.
06 — When to act
A protocol that produces a clean upward trend over 90 days, with each point above the previous one allowing for noise, is working. Don't change it. A protocol that produces a flat line for 90 days is not working, regardless of what the number is.
07 — Closing
The point of biomarker tracking isn't to react to numbers. It's to see patterns. Patience pays here in a way it rarely does in the rest of self-experimentation.