Cortisol Is a Curve: Why One Hormone Test Can Mislead You
Marina Rivas, cofounder and CEO of Eli Health, on what hormone trends can show, what a single lab misses, and why more testing is not automatically better.
Published June 10, 2026
A single morning cortisol or hormone lab is a snapshot of a moving target. It can miss timing, wake time, sleep, stress, training load, cycle context, and recovery. Marina Rivas, cofounder and CEO of Eli Health, is building at-home saliva testing to capture hormones as patterns over time, and in this conversation we pressure test what that approach can and cannot tell you.
We get specific about cortisol curves versus cortisol panic, what saliva testing actually measures, progesterone and testosterone tracking, and how to tell a trend from noise. We also draw the line that matters: wellness hormone trend data is not diagnosis or treatment. Symptoms, abnormal results, HRT or TRT decisions, adrenal concerns, and medication or supplement changes belong with a qualified clinician.
Before you watch
- Hormones like cortisol move on a daily curve, so one timed lab can mislead unless you know wake time, sleep, and context around it.
- Trends over time are more informative than any single reading, but a trend is only useful once you can separate real signal from day-to-day noise.
- Saliva testing can surface patterns for wellness tracking, but it is not definitive and does not diagnose or rule out a condition.
- More testing is not automatically better. Symptoms, abnormal results, and any HRT, TRT, adrenal, medication, or supplement decision belong with a clinician.
Chapters
00:00
Hormones are patterns, not snapshots
Why a single reading rarely tells the whole story.
00:54
Why annual hormone testing misses context
One yearly lab drops the timing and lifestyle context around it.
02:43
How at-home saliva testing works
What the method measures and how samples are collected.
06:17
Cortisol curves vs cortisol panic
Reading the shape of the day instead of reacting to one number.
12:14
Why timing and baseline matter
The same value means different things depending on when it is taken.
15:05
Stress, training, and overtraining signals
How load and recovery show up in the data, and their limits.
18:43
Progesterone and testosterone tracking
What tracking these hormones over time can and cannot show.
20:40
Trends vs noise in hormone data
Telling a real pattern apart from normal variation.
23:44
Designing testing people can actually use
Why usability shapes whether the data is worth anything.
24:40
Cortisol, performance, and recovery
How cortisol relates to training and rest, with caveats.
28:31
What changes testosterone levels?
Factors that move the number, and why context matters.
31:36
Wellness data vs clinical diagnostics
Where consumer tracking ends and clinical testing begins.
35:18
AI needs better biological inputs
Why models in health depend on the quality of the data underneath.
37:39
Why health hardware takes years to build
The slow work of validating at-home testing.
39:04
What longevity-minded people should take away
How to use hormone trends without overreading them.