The epidemiology of glioma in adults: a “state of the science” review

This is a wide-ranging and comprehensive study.¹ However, the section “Nonionizing Radiation: Cellular Phones” has serious deficiencies. It cites 3 incidence time trend studies,^2–4 2 cohort studies,^5,6 and 1 case control study.⁷

Incidence Time Trend Studies

Late ascertainment and poor histological concordance are common accuracy problems. Late ascertainment results in an underestimation of incidence rates in recently reported years. A paper reported: “Results: Initial incidence case counts … accounted for only 88%–97% of … final counts; it would take 4–17 years for 99% or more of the cancer cases to be reported.”⁸ Another study reported that the histological concordance by 4 neuropathologists reviewing gliomas was “52% all 4 reviewers, any 3 reviewers, 60%; 2 reviewers, 70%.”⁹

The study by Deltour and colleagues² reported that glioma rates were stable among the 40–59 age group from 1979 to 2008.² The Ostrom authors¹ failed to report a significant increase, ages 20–79, annual percent change (APC) = 0.4%, 95% CI = 0.1%–0.6% in men and APC = 0.3%, 95% CI = 0.1%–0.5% in women.² It received funding via a “firewall” from the cellphone companies Telia-Sonera, Telenor, and Ericsson.¹⁰ During the years of this study (1979 to ∼1994; 53% of the duration), cellphones did not exist or the prevalence was very low; in 1998 the prevalence was 44%; by 2005 prevalence had reached 100%. With incidence trends over a 30-year period where in most of the years there was almost no cellphone use and with only 3 years of 100% prevalence, how can one conclude whether or not cellphone use was affecting incidence?

The US study by Little and colleagues³ was for the years 1992–2008.³ In 1992 cellphone prevalence was 1% and by 2008 it was 84%.¹¹ A 2013 report noted that the Veterans Administration hospitals had ceased from 2005 to 2014 to report cancer cases diagnosed among military veterans.¹² The result was that 3%–8% of all male cancer cases were missing. In 1992, only 1% of the population were using cellphones, whereas by 2008, use was at 84%. With 3%–8% of male cancers not reported, combined with late ascertainment, how could a change in glioma incidence rates be expected? In spite of these issues, Little et al reported a significantly increased APC in temporal lobe glioma, APC = 0.73%, 95% CI = 0.23%–1.23%. The temporal lobe absorbs the largest proportion of cellphone radiation of any anatomic region of the brain.¹³

The title of the third incidence time trend study cited, “Changes in Brain Glioma Incidence and Laterality Correlates With Use of Mobile Phones—a Nationwide Population Based Study” (emphasis added),⁴ is in direct contradiction to the assertions in the deficient section.

Incidence Time Trend Studies not Cited

A US paper examined cancer incidence across 3 cancer registries for the years 1992–2006.¹⁴ It reported: “Data from 3 major cancer registries demonstrate increased [APC] incidences of GBMs in the frontal lobe, temporal lobe, and cerebellum.” These 3 anatomic regions absorb between 81% (900 MHz) and 86% (1800 MHz) of all the cellphone radiation absorbed by the brain.¹³

An Australian paper with 2000–2008 data,¹⁵ though cited in the Ostrom study,¹ was not cited for its time trend results, brain cancer APC = 3.9%, 95% CI = 2.4–5.5.¹⁵ The same team reported: “A significant increasing incidence in glioblastoma multiforme … was observed in the study period [APC = 2.5%, 95% CI = 0.4–4.6], particularly after 2006.”¹⁶

For the years 2003–2012 the Danish Cancer Registry reported an increased incidence of male and female brain cancers of 41.2% and 46.1%, respectively.¹⁷

Cohort Studies

The Ostrom study cited 2 cohort studies as evidence that cellphone use is not a risk for glioma.^5,6 For rare diseases, case control studies are essential. Cohort studies are incapable of determining risks.¹⁸ It is axiomatic that absence of evidence is not evidence of absence. Both studies found significant reduced risks for various cancers.

Case Control Study

A single case control study was cited, noting that its odds ratios (ORs) “were markedly elevated in all categories of use.”⁷

Case Control Studies not Cited

The Hardell team's significant findings are consistent with what would be expected if wireless phones (cell and cordless) were causing brain cancer:

1. The higher the cumulative hours of use, the higher the risk.^7,19

2. The longer the time since first use, the higher the risk.^7,19–21

3. The higher the radiated power, the higher the risk.^7,19–21

4. Ipsilateral risk is higher than contralateral risk.^7,19,20,22

Another study reported brain cancer risk with 1640+ hours of cellphone use compared with <5 hours of use, OR = 1.82, 95% CI = 1.15–2.89; and for 10+ years since use compared with 1–1.9 years since first use, OR = 2.18, 95% CI = 1.43–3.31.²³

The CERENAT study reported “heavy mobile phone use” (≥896 hours), OR = 2.89, 95% CI = 1.41–5.93; with 5+ years since first use, OR = 5.30, 95% CI = 2.12–13.23, P<.001; and for use exclusively in urban areas, OR = 8.20, 95% CI = 1.37–49.07.²⁴

In conclusion, the authors' statement that the “evidence published since the IARC monograph in 2011 does not support an association between cellular phone use and the risk of glioma in adults¹” requires revision.