The WHO-commissioned review on phones and brain cancer, and how it compares
The review by Ken Karipidis and ten co-authors, commissioned by the World Health Organization and published in Environment International in 2024, pooled 63 studies from 1994 to 2022. It found no increased risk of glioma among people who used mobile phones, with a pooled relative risk of 1.01, and rated its certainty in that result as moderate. That reads differently from IARC's 2011 label of "possibly carcinogenic," and IARC's advisers have asked for a new evaluation before 2030, though none is on its meeting calendar yet.
What the review set out to do
The paper is Part I of the authors' review, covering the most-studied tumors, and one of several systematic reviews WHO commissioned on radio-frequency fields. WHO paid for part of it, with co-funding from New Zealand's Ministry of Health, Italy's national health institute and ARPANSA, Australia's radiation protection agency, where the lead author works. The protocol was registered in advance (PROSPERO CRD42021236798) and published in 2021, so the methods were public before the analysis.
The team included only cohort and case-control studies, and grouped exposure three ways:
- phones held against the head
- whole-body exposure from fixed transmitters such as broadcast antennas and cell towers
- work with hand-held radios or other radio-frequency equipment
Each study was scored for risk of bias with a tailored version of a tool called OHAT. Confidence in each conclusion was then graded with the GRADE approach, which is where the "moderate" and "low" certainty labels below come from. In all, the 63 papers covered participants in 22 countries and 119 combinations of exposure and tumor type.
What it found
For people who had ever used a mobile phone, or used one regularly, compared with people who hadn't, the pooled risks clustered around 1.0. A value of 1.0 means no difference.
Every confidence interval includes 1.0, so none of these differs from no effect at the usual statistical threshold.
| Tumor type | Pooled relative risk (95% CI) |
|---|---|
| Glioma | 1.01 (0.89 to 1.13) |
| Meningioma | 0.92 (0.82 to 1.02) |
| Acoustic neuroma | 1.03 (0.85 to 1.24) |
| Pituitary tumors | 0.81 (0.61 to 1.06) |
| Salivary gland tumors | 0.91 (0.78 to 1.06) |
| Brain tumors in children and young adults | 1.06 (0.74 to 1.51) |
Source: Karipidis et al., Environment International 191:108983 (2024), abstract. Dashed line: a relative risk of 1.0, meaning no difference.
The review also checked whether risk rose with more exposure. It found no increase for glioma, meningioma or acoustic neuroma with longer time since first use, more total call time or more calls. Among the studies at lowest risk of bias, people with 10 or more years of use had a pooled risk of 0.95 for glioma and 1.00 for acoustic neuroma.
The certainty ratings matter as much as the numbers. Here is how the review graded each line of evidence:
- Mobile phones and the six tumor types above: moderate certainty that phone use likely does not raise the risk.
- Cordless home phones: low certainty that they may not raise the risk of glioma, meningioma or acoustic neuroma.
- Broadcast antennas and cell towers: moderate certainty of no increase in childhood leukemia, low certainty for brain tumors in children.
- Workplace exposure: low certainty of no increase in brain cancer or glioma.
How it compares with IARC's 2011 rating
The two are answering different questions. IARC Monographs identify factors that can increase the risk of cancer and grade how strong the evidence is, without estimating how large the risk would be. In May 2011 a working group of 31 scientists from 14 countries placed radio-frequency fields in Group 2B, "possibly carcinogenic to humans" (IARC press release). The full Monograph Volume 102 followed in 2013.
The rating rested on "limited" human evidence for glioma and acoustic neuroma. IARC defines limited as a positive association with a credible causal reading where chance, bias or confounding could not be ruled out. The group did not put a number on the risk, though its release pointed to one study of phone use up to 2004 in which the heaviest users had a 40% higher risk of glioma. IARC also says its group labels "have no quantitative significance." A 2B listing describes how strong the evidence is, not how big a risk would be.
The 2024 review had more to work with. It took in cohort results published after 2011, used bias ratings and asked how confident anyone can be. In 2024, IARC's advisers also cited a bias analysis that found the Interphone pattern compatible with no real effect, if people with tumors misreported their phone use more than the controls did.
The big studies side by side
| Study | Design | Size | Main finding |
|---|---|---|---|
| IARC Monograph 102 (2011 meeting) | Expert hazard evaluation | 31 scientists, 14 countries | Group 2B; limited evidence for glioma and acoustic neuroma |
| Interphone (2010) | Case-control, 13 countries | 2,708 glioma and 2,409 meningioma cases | No overall increase (glioma odds ratio 0.81 for regular users); 1.40 in the top tenth of call time, 1,640 hours or more |
| Danish cohort (2011) | Nationwide cohort, subscription records | 358,403 subscribers, 3.8 million person-years | No increase; 13+ years of subscription gave a rate ratio of 1.03 in men |
| Million Women Study (2022) | Prospective cohort, UK | 776,156 women, 14 years, 3,268 brain tumors | Relative risk 0.97 for all brain tumors, 0.89 for glioma |
| COSMOS (2024) | Prospective cohort, 5 countries | 264,574 people, median 7.12 years | Hazard ratio 1.00 for glioma per 100 hours of calls |
| Karipidis et al. (2024) | Systematic review of human studies | 63 papers | Pooled glioma risk 1.01; moderate certainty |
| Mevissen et al. (2025) | Systematic review of animal studies | 51 studies after a 2026 correction | High certainty for glioma and heart schwannoma in male rats |
Each study had a known weak spot. Interphone's authors wrote that "biases and error prevent a causal interpretation," and they flagged implausible reported call times in the top group. The Danish study used subscription records, and the National Cancer Institute notes that a subscriber isn't always the person using the phone. IARC's advisers called the Million Women exposure data crude. COSMOS had a median follow-up of just over seven years.
Where the review has critics
In early 2025, seven critics published a letter in Environment International titled "The Systematic Review on RF-EMF Exposure and Cancer by Karipidis et al. (2024) has Serious Flaws that Undermine the Validity of the Study's Conclusions" (letter). The signers included Hardell, whose team published the 2011 Swedish case-control study discussed in our registry article, and Melnick, a retired NTP scientist. The review's authors published a reply in the same issue.
The paper is open about its own limits. It found one study with an outsized pull on the glioma result; leaving it out moved the pooled risk to 0.96. It also said the size and direction of the biases in the underlying studies "proved difficult to predict." A meta-analysis can only be as good as the studies it pools. As the FDA has pointed out, studies outside the lab have no direct measurement of radio-frequency exposure and rely mostly on what participants report.
The sharpest contrast comes from WHO's own series. The companion review of animal studies, led by Meike Mevissen and published in May 2025, rated the evidence as high certainty for gliomas and heart schwannomas in male rats, based on two chronic bioassays. Its authors noted that these are the same two tumor types IARC flagged in humans. They also wrote that extrapolating from rodents to people is "particularly complex" for radio-frequency fields. A correction in 2026 revised its study counts but kept those ratings. Since then, two matched rat studies from Japan and Korea, published in 2026, found no tumor increase at 4 W/kg; our NTP explainer covers them.
Has IARC scheduled a re-evaluation?
Not yet. An IARC advisory group of 28 scientists from 22 countries met in March 2024 to set priorities for 2025 to 2029. Its summary in The Lancet Oncology lists radio-frequency radiation as high priority and advises doing it in the latter half of the five-year period (summary). The full advisory report gives the reasoning: new human and animal evidence "could support a re-evaluation, although a change in the current classification of the carcinogenicity of RF-EMF is uncertain," and the group wanted to wait for ongoing cancer bioassays.
As of November 2026 the Monographs meetings IARC has announced run through Volume 144, on lead and cadmium, set for October 2027 (meeting page). None of the announced volumes covers radio-frequency fields. WHO's separate health risk assessment, meant to become an Environmental Health Criteria monograph that updates its 1993 one, is still described as in progress on the WHO project page.
What this means if you're deciding something
None of these studies tested consumer products. Shops sell accessories to people following this debate; the EMFCALM.com shop, for example, sells EMF stickers for phones and other devices, and EMFCALM's packs come in sets of 3, 6 and 12. The evidence on this page is about phones and tumors, and it says nothing either way about any product. Personal health questions, especially about symptoms or a family history of brain tumors, are for your doctor.
Whenever IARC meets, its working group will have three sets of results in front of it: human studies that WHO's reviewers rated moderate certainty for no effect, animal studies that WHO's other reviewers rated high certainty for an effect in male rats, and two 2026 rat studies that found no tumor increase at 4 W/kg. IARC's advisers have already written that a change to the 2B label is uncertain.