This is part of our Study Spotlight series, where we break down the latest peer-reviewed EMF research into plain language. No hype, no dismissal — just what the science actually says.
If you’ve spent any time reading about 5G health concerns online, you’ve probably encountered claims about millimeter waves — the higher-frequency bands that represent the truly new part of 5G technology. Specifically, the 26 GHz band (sometimes called “mmWave 5G”) is one that has almost no track record in telecommunications, making it the biggest unknown in the 5G rollout.
A team of French researchers just published one of the first controlled human studies testing what actually happens when you expose people to this frequency. Their finding: nothing measurable.
Why This Study Matters
Here’s the thing about 26 GHz: unlike the sub-6 GHz frequencies used by most “5G” networks today (which are basically souped-up 4G), 26 GHz millimeter waves are genuinely new territory for widespread civilian exposure. Previous generations of wireless technology never operated at these frequencies.
That novelty is exactly what makes people nervous — and exactly why this study from Lisa Michelant and colleagues at INERIS (France’s national institute for environmental safety) is so important. It’s among the first controlled human exposure experiments at this specific frequency.
If you’ve been following our Study Spotlight series, you might recognize INERIS — the same lab that ran the first controlled 5G human exposure study at 3.5 GHz (our Spotlight #6). Now they’ve moved up the frequency spectrum to test the band that generates the most public concern.
The Setup
This study was triple-blind — meaning the participants, the researchers administering the exposure, and the people analyzing the samples all had no idea who received real vs. sham exposure. That’s about as rigorous as study design gets.
31 healthy adults completed the main protocol:
| Parameter | Detail |
|---|---|
| Frequency | 26 GHz (millimeter wave 5G band) |
| Duration | 26.5 minutes of continuous exposure |
| Power density | 2 V/m at the head, 1 V/m at the torso |
| Design | Triple-blind, randomized, sham-controlled |
| Participants | 31 adults (main protocol), 16 (exploratory protocol) |
The exposure levels were chosen to match upper-range real-world environmental measurements — meaning the levels you might actually encounter near a 5G mmWave base station, not some artificially jacked-up laboratory dose.
Each participant went through both conditions (real exposure and sham) on separate days, so they served as their own control. EEG (brain wave) recordings were taken simultaneously.
Check your EMF exposure
See cell towers, power lines, and substations near any US address.
Search Your AddressWhat They Measured
The researchers focused on two well-established stress biomarkers collected through saliva samples:
Salivary cortisol — the body’s primary stress hormone. If 5G exposure triggered a physiological stress response, cortisol would be one of the first things to spike. It’s the same hormone that surges when you’re anxious, sleep-deprived, or under threat.
Salivary alpha-amylase — a marker of autonomic nervous system activation, specifically the sympathetic (“fight or flight”) branch. This enzyme is released by salivary glands in response to sympathetic nerve stimulation and responds faster than cortisol, making it a good acute stress indicator.
Saliva was collected at multiple time points: before, during, and after exposure in both protocols.
The exploratory protocol (16 participants) added cognitive testing and higher temporal sampling resolution — saliva collected every 5 minutes — to catch any transient effects that might be missed with less frequent sampling.
The Results
No significant differences between real and sham exposure. For either biomarker. At any time point.
| Biomarker | Real Exposure | Sham Exposure | Significant Difference? |
|---|---|---|---|
| Cortisol | Normal temporal pattern | Normal temporal pattern | ❌ No |
| Alpha-amylase | Normal temporal pattern | Normal temporal pattern | ❌ No |
Both biomarkers showed their expected natural fluctuations throughout the experimental session (cortisol tends to decline gradually over the morning, for example), but these patterns were identical whether the 5G transmitter was on or off.
The high-resolution exploratory protocol with 5-minute sampling intervals confirmed the same finding — consistent temporal stability in both conditions. Even with the fine-grained measurement that would catch a brief spike, nothing appeared.
Why This Is Significant
Several things make this study particularly noteworthy:
1. The Frequency Actually Matters
Most existing RF-EMF health studies test frequencies between 900 MHz and 3.5 GHz. The 26 GHz band behaves very differently — it’s absorbed almost entirely by the skin (penetration depth of about 1mm) rather than passing into deeper tissues. This means the biological interaction mechanism is fundamentally different from lower-frequency radio waves.
Some researchers have hypothesized that millimeter waves could trigger stress responses through skin nerve endings or sweat duct interactions, even if they don’t penetrate deep. This study tested that hypothesis directly — and didn’t find evidence for it.
2. Triple-Blind Design Eliminates Nocebo
One of the biggest challenges in RF-EMF research is the nocebo effect — people who believe they’re being exposed to harmful radiation can actually develop real symptoms from the belief alone. Triple-blinding is the gold standard for eliminating this confound, and this study achieves it.
3. Realistic Exposure Levels
The 2 V/m exposure level wasn’t arbitrary. It was calibrated to match the upper end of what people might actually encounter near 5G mmWave installations. This matters because laboratory studies often use exposure levels far above real-world conditions, making their results less applicable to everyday life.
4. Within-Subject Design
Each person served as their own control, experiencing both real and sham exposure. This eliminates between-person variability (genetics, stress sensitivity, lifestyle) as a confounding factor.
The Caveats
As always, we need to be honest about what this study doesn’t tell us:
Short-term only. This tested 26.5 minutes of exposure. It can’t speak to what happens with months or years of chronic exposure — which is ultimately what matters for people living near 5G base stations.
Small sample. 31 participants (16 in the exploratory arm) provides reasonable statistical power for large effects but might miss subtle biological changes. Larger studies are needed.
Two biomarkers. Cortisol and alpha-amylase are established stress markers, but the human stress response involves dozens of hormones, neurotransmitters, and immune mediators. A broader biomarker panel could reveal effects not captured here.
Healthy adults only. The researchers specifically note that “further research is warranted to evaluate… potential impacts in vulnerable populations.” Children, elderly people, and those with existing health conditions might respond differently.
Skin-deep exposure. At 26 GHz, nearly all the energy is absorbed by the skin. This study measured systemic stress markers, which is appropriate for a general stress response, but local skin effects (inflammation, nerve activation) weren’t assessed. A separate 2025 study at 3.5 GHz found no oxidative stress or DNA repair impairment in human skin fibroblasts even at 4 W/kg.
How This Connects
This result aligns with a growing body of evidence from controlled human 5G studies:
- The INERIS 3.5 GHz study (same research group, lower frequency) also found no clinically significant effects on stress markers or cognitive function
- The 5G phone SAR brain simulation study showed peak brain exposure at just 1.7% of safety limits
- The operating room RF study found occupational RF exposure well below 1% of safety limits
At the same time, it’s worth remembering that other research has found biological effects at RF frequencies — the cell tower proximity blood study found immune cell changes in people living very close to towers, and the ICBE-EMF analysis argues that current safety limits may be too permissive. The evidence picture remains complex.
The Bottom Line
This study provides reassuring evidence that acute exposure to 26 GHz 5G millimeter waves, at levels matching real-world conditions, does not trigger a measurable stress response in healthy adults. The triple-blind design and within-subject controls make this one of the more methodologically rigorous studies in the field.
However, it’s a single acute-exposure experiment with a limited number of participants and biomarkers. The real public health question — whether chronic, long-term exposure to these frequencies has cumulative effects — remains open and will require longitudinal studies to answer.
For now, if you’re worried specifically about the “millimeter wave” aspect of 5G, this is one data point suggesting it may be less alarming than feared. But “one study found nothing in 26 minutes” is very different from “decades of exposure is definitely safe.” Science is a slow process, and honest researchers — including the authors of this study — will be the first to tell you that.
Want to check what wireless infrastructure exists near you? Use our EMF Radar map tool to see cell towers in your area, including any 5G installations.
Study: Michelant L, Hugueville L, Lévêque P, Selmaoui B. “No measurable impact of acute 26 GHz 5G exposure on salivary stress markers in healthy adults.” Environmental Research, 290:123439, February 2026. PubMed | DOI
Related Reading
- Study Spotlight: Italian Researchers Blasted Brain Cancer Cells with 4G LTE and WiFi Simultaneously — Nothing Happened
- Study Spotlight: Could the Same Radio Waves From Your Phone Help Treat Alzheimer’s?
- Study Spotlight: Living Near Cell Towers Changed These People’s Blood — Here’s What Researchers Found
- Study Spotlight: Radiofrequency Radiation Actually Helped Blood Stem Cells — Here’s How
Concerned about EMF? Check your address on EMF Radar to see nearby towers and power lines, or find a certified EMF consultant for professional testing.