Exposure to radiofrequency radiation increases the risk of breast cancer: A systematic review and meta-analysis
Shih YW, O'Brien AP, Hung CS, Chen KH, Hou WH, Tsai HT.
- DOI
- 10.3892/etm.2020.9455
- Record issued
- 2026-08-05
- Engine
- 7.15.0
- Exported
- 2026-09-19
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/1dd8c7b7-d012-4e22-9e65-4777a5993ea8 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern−0.5★
- ReportingBiological variables partially met−0.25★
- No data or code availability links were detected to verify.
- 01Methods and results do not match
The Newcastle-Ottawa Scale (NOS) table (Table II) is labeled 'Case control study' but includes cohort studies (Tynes et al, 1996; Kliukiene et al, 1999; Pollán et al, 2001) without adapting the scale for cohort design. This may affect the validity of the quality assessment.
A quality assessment method for case and control studies was developed based on the Newcastle-Ottawa Scale (NOS).
Methodsreviewer’s wording
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This meta-analysis is methodologically sound with a registered protocol, clear inclusion criteria, appropriate statistical tests, and transparent reporting. The principal weakness is the incomplete reporting of participant-level biological variables (age, sex, health status) across the included studies, which limits the ability to assess potential confounding or effect modification.
This meta-analysis of eight observational studies was evaluated across eight rigor dimensions. Several sub-criteria within dimensions (e.g., randomization, blinding, power analysis) were not applicable due to the meta-analysis design. The three independent reviewers largely converged on all dimensions; minor disagreements on sub-criteria (e.g., power analysis, exact p_values) were resolved by weighing the evidence and the paper's methodology.
Numerical inconsistencies
None foundValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
Checked — nothing surfaced.
Recomputed 3 tests: 3 consistent, 0 inconsistent; 3 via agent-written checks.
- CONSISTENTreported p = .014 · recomputed p = .014Reviewers 1, 2Heterogeneity Q-testHow we recomputed it: pChi2(17.6, 7)
- CONSISTENTreported p = .048 · recomputed p = .049Reviewers 1, 2Egger's test for publication biasHow we recomputed it: 2*(1-tCdf(2.46, 6))
- CONSISTENTreported p = .040 · recomputed p = .042Reviewer 2Sensitivity analysis heterogeneity Q-test: Q=13.04, df=6, paper reports P=0.04How we recomputed it: pChi2(13.04, 6)
Overstated conclusions
None foundConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
Checked — nothing surfaced.
11 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewers 1, 2, 3Radiofrequency radiation exposure significantly increases the risk of breast cancer.The pooled RR of 1.189 (95% CI 1.056-1.339) from the meta-analysis supports this claim.Evidence: Pooled RR=1.189; 95% CI, 1.056-1.339
“The association between radiofrequency radiation exposure and the risk of breast cancer was significant (; pooled RR=1.189; 95% CI, 1.056-1.339).”
Results - supportedReviewer 1The risk is especially increased in women aged ≥50 years.Subgroup analysis shows RR=2.179 (95% CI 1.260-3.770) for this age group.Evidence: RR=2.179; 95% CI, 1.260-3.770
“The results indicated that radiofrequency radiation exposure significantly increased the risk of breast cancer susceptibility among subjects aged ≥50 years (; RR=2.179; 95% CI, 1.260-3.770)”
Results - supportedReviewer 1Use of mobile phones and computers significantly increases breast cancer development.Subgroup analysis shows RR=2.057 (95% CI 1.272-3.327) for electric appliance use.Evidence: RR=2.057; 95% CI, 1.272-3.327
“mobile phone/computer exposure significantly increased the risk of breast cancer (; RR=2.057; 95% CI, 1.272-3.327)”
Results - supportedReviewer 1Occupational radiofrequency exposure and transmitter exposure do not increase breast cancer risk.The RRs for occupational (1.274, 95% CI 0.956-1.697) and transmitter (1.133, 95% CI 0.987-1.300) exposures are not statistically significant.Evidence: RR=1.274; 95% CI, 0.956-1.697 and RR=1.133; 95% CI, 0.987-1.300
“a significant association was not observed for radiofrequency radiation exposure in an occupational environment (; RR=1.274; 95% CI, 0.956-1.697) or for transmitter exposure (; RR=1.133; 95% CI, 0.987-1.300)”
Results - supportedReviewers 1, 2, 3This is the first meta-analysis on radiofrequency radiation and breast cancer risk.The authors state this based on their literature search; no contradictory evidence is presented.Evidence: Statement of novelty in introduction
“to the best of our knowledge, the present study performed the first meta-analysis aiming to evaluate and obtain more precise and comprehensive estimates of the association between radiofrequency radiation exposure and the risk of breast cancer.”
Introduction - supportedReviewer 2Radiofrequency radiation exposure significantly increases breast cancer risk among women aged ≥50 years.The subgroup analysis shows RR=2.179 (95% CI 1.260-3.770) for age ≥50, supporting the claim.Evidence: RR=2.179; 95% CI, 1.260-3.770
“Subgroup analyses indicated that radiofrequency radiation exposure significantly increased the risk of breast cancer susceptibility among subjects aged ≥50 years (RR=2.179; 95% CI, 1.260-3.770).”
Abstract - supportedReviewer 2Use of mobile phones and computers significantly increases breast cancer risk.The subgroup analysis shows RR=2.057 (95% CI 1.272-3.327) for electrical appliance use, supporting the claim.Evidence: RR=2.057; 95% CI, 1.272-3.327
Pooled estimates revealed that the use of electrical appliances, which emit radiofrequency radiation, such as mobile phones and computers, significantly increased breast cancer development (RR=2.057; 95% CI, 1.272-3.327).
Abstractreviewer’s wording - supportedReviewer 2Occupational radiofrequency exposure does not significantly increase breast cancer risk.The subgroup analysis shows RR=1.274 (95% CI 0.956-1.697), which is not significant, supporting the claim.Evidence: RR=1.274; 95% CI, 0.956-1.697
“occupational radiofrequency exposure and transmitters did not increase breast cancer development (RR=1.274; 95% CI, 0.956-1.697; RR=1.133; 95% CI, 0.987-1.300, respectively)”
Abstract - supportedReviewer 3Radiofrequency radiation exposure significantly increases breast cancer risk in women aged ≥50 years.Subgroup analysis shows RR=2.179 (95% CI 1.260-3.770), supporting the claim.Evidence: Subgroup analysis: RR=2.179; 95% CI, 1.260-3.770.
“radiofrequency radiation exposure significantly increased the risk of breast cancer susceptibility among subjects aged ≥50 years (RR=2.179; 95% CI, 1.260-3.770).”
Results - supportedReviewer 3Use of electrical appliances (mobile phones, computers) emitting radiofrequency radiation significantly increases breast cancer risk.Subgroup analysis shows RR=2.057 (95% CI 1.272-3.327), supporting the claim.Evidence: Subgroup analysis: RR=2.057; 95% CI, 1.272-3.327.
Pooled estimates revealed that the use of electrical appliances, which emit radiofrequency radiation, such as mobile phones and computers, significantly increased breast cancer development (RR=2.057; 95% CI, 1.272-3.327).
Abstractreviewer’s wording - supportedReviewer 3Occupational radiofrequency exposure and transmitter exposure do not significantly increase breast cancer risk.Subgroup analyses show non-significant associations (RR=1.274 and 1.133, with CIs crossing 1.0), supporting the claim.Evidence: Occupational: RR=1.274; 95% CI, 0.956-1.697; Transmitter: RR=1.133; 95% CI, 0.987-1.300.
“occupational radiofrequency exposure and transmitters did not increase breast cancer development (RR=1.274; 95% CI, 0.956-1.697; RR=1.133; 95% CI, 0.987-1.300, respectively).”
Abstract
Data authenticity concerns
1 finding · worst mediumAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
- Methods and results do not matchAssessed
1 integrity concern flagged (0 high).
- mediummethod result mismatchThe Newcastle-Ottawa Scale (NOS) table (Table II) is labeled 'Case control study' but includes cohort studies (Tynes et al, 1996; Kliukiene et al, 1999; Pollán et al, 2001) without adapting the scale for cohort design. This may affect the validity of the quality assessment.
A quality assessment method for case and control studies was developed based on the Newcastle-Ottawa Scale (NOS).
Methodsreviewer’s wording
Reporting gaps
1 finding · worst mediumRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Biological variables underreported (sex, age, strain)Assessed
The Introduction cites relevant literature on radiofrequency and cancer, including IARC classification, and identifies a gap in the literature regarding a meta-analysis on radiofrequency radiation and breast cancer specifically. The hypothesis follows from the cited evidence. However, limitations of prior reviews are not explicitly addressed.
“Previous studies have also focused on the effects of the exposure to non-ionizing radiofrequencies on brain tumors, leukemia, salivary gland tumors, infertility and electro-hypersensitivity”
“it was hypothesized that exposure to radiofrequency radiation may induce breast cancer development due to the induction of oxidative stress and apoptosis in breast cancer cells.”
“to the best of our knowledge, the present study performed the first meta-analysis aiming to evaluate and obtain more precise and comprehensive estimates of the association between radiofrequency radiation exposure and the risk of breast cancer.”
“Radiofrequency radiation, which is a subcategory of non-ionizing radiation, has been indicated to exhibit harmful effects that are similar to those of ionizing radiation, and to increase the risk of cancer ().”
“the majority of meta-analysis studies have focused on the association between mobile phones and tumors () or electromagnetic fields and breast cancer ().”
“To the best of our knowledge, the present study performed the first meta-analysis aiming to evaluate and obtain more precise and comprehensive estimates of the association between radiofrequency radiation exposure and the risk of breast cancer.”
The paper reports a detailed search strategy across nine databases, PRISMA flow diagram, inclusion/exclusion criteria, and quality assessment using the Newcastle-Ottawa Scale. While no formal power analysis was performed, search completeness is demonstrated by the extensive search. Heterogeneity was assessed, and a sensitivity analysis was conducted. The study registration with PROSPERO is provided.
“a sensitivity analysis was conducted via deleting one study to examine the influence of individual datasets”
“For inclusion, the studies were required to meet all the following criteria: i) Evaluated associations between radiofrequency radiation and the risk of breast cancer; ii) studied a human population; iii) provided detailed data for calculating the relative risk (RR) or odds ratio (OR) and 95% confidence interval (CI); and iv) investigated radiofrequency radiation or any frequency classified as radiofrequency.”
“sensitivity analysis was conducted by removing one study () from the analysis to detect the pooled RR estimates”
“For inclusion, the studies were required to meet all the following criteria: i) Evaluated associations between radiofrequency radiation and the risk of breast cancer; ii) studied a human population; iii) provided detailed data for calculating the relative risk (RR) or odds ratio (OR) and 95% confidence interval (CI); and iv) investigated radiofrequency radiation or any frequency classified as radiofrequency.”
The included studies are on breast cancer, so the population is assumed female, but sex is not explicitly stated for all studies. Age is partially reported in subgroup analyses but not in the study characteristics table. No information on health status, race/ethnicity, or comorbidities is provided. The paper does not meet the 60% adequacy threshold for applicable biological variables.
“radiofrequency radiation exposure significantly increased the risk of breast cancer, especially in women aged ≥50 years”
“Exposure to radiofrequency radiation increases the risk of breast cancer: A systematic review and meta-analysis”
“A total of three studies evaluated an age group of ≥50 years old.”
The paper states 'Ethics approval and consent to participate Not applicable.' The study does not involve new subject data.
“Ethics approval and consent to participate Not applicable.”
“Not applicable.”
“Not applicable.”
The paper uses Comprehensive Meta-Analysis v2.0 (Biostats, Inc.) and identifies it by name and version. No other biological or chemical resources are used.
“All data analyses were performed with Comprehensive Meta-Analysis v2.0 software (Biostats, Inc.).”
“All data analyses were performed with Comprehensive Meta-Analysis v2.0 software (Biostats, Inc.).”
“All data analyses were performed with Comprehensive Meta-Analysis v2.0 software (Biostats, Inc.).”
The Cochran Q-test and I² are used to assess heterogeneity, and Egger's test and trim-and-fill for publication bias. The overall and subgroup results are presented as RR with 95% CI. The software is identified. The paper does not report exact p-values for the main effect, but this is not required for a meta-analysis that relies on effect estimates and CIs.
“pooled RR=1.189; 95% CI, 1.056-1.339”
“Comprehensive Meta-Analysis v2.0 software (Biostats, Inc.)”
“pooled RR=1.189; 95% CI, 1.056-1.339”
“All data analyses were performed with Comprehensive Meta-Analysis v2.0 software (Biostats, Inc.).”
“pooled RR=1.189; 95% CI, 1.056-1.339”
The paper states 'All data generated or analyzed during this study are included in this published article.' No repository deposit is needed as the data are presented in tables. No custom code was used.
“All data generated or analyzed during this study are included in this published article.”
“All data generated or analyzed during this study are included in this published article.”
“All data generated or analyzed during this study are included in this published article.”
The paper is registered with PROSPERO, follows PRISMA guidelines, reports all outcomes, discusses limitations (sample size, dose-response, publication bias), and provides funding and competing interests.
“registered with the PROSPERO database (registration no. CRD42018087283)”
“In accordance with Preferred Reporting Items for Systematic Reviews and Meta-analysis, an evaluation protocol was prepared and registered with the PROSPERO database”
“registered with the PROSPERO database (registration no. CRD42018087283).”
“The current study presents certain limitations. Firstly, only results in the selected papers were used, which limited the analyses.”
“an evaluation protocol was prepared and registered with the PROSPERO database (registration no. CRD42018087283).”
“The Preferred Reporting Items for Systematic Reviews and Meta-Analyses () flow diagram of the review process is presented in .”
“The current study presents certain limitations. Firstly, only results in the selected papers were used, which limited the analyses.”
Registered (1 ID: PROSPERO). No reporting guideline cited.
Broken references and links
None found · partly checkedReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Nothing surfaced — but not everything feeding this category ran (missing: data/code link verification), so read this as a partial clean bill.
Checked 57 references by DOI: 52 verified — 5 no DOI (shown, not verified).
- NO DOIEffect of exposure to 900 MHz GSM mobile phone radiofrequency radiation on estrogen receptor methylation status in colon cells of male sprague dawley ratsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOINon-ionizing radiation, Part 2: Radiofrequency electromagnetic fieldsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIPrinciples of biostatistics (2nd edition)No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIHousehold electromagnetic fields and breast cancer in elderly womenNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIThe effect of current environmental risk factors on breast cancerNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
Copyediting
9 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 9 minor suggestions below.
9 copyedit issues flagged: mostly consistency, grammar, other.
- MINORgrammarAbstract“A total of eight studies (four case-control and four cohort studies) were eligible for quantitative analysis.”→ Consider removing 'A total of' for conciseness: 'Eight studies (four case-control and four cohort) were eligible for quantitative analysis.'Repetitive phrasing throughout the manuscript.
- MINORconsistencyTable I header“No. cases/Total population”→ Ensure consistent use of 'No.' vs 'n' (e.g., 'n cases/Total population')Minor formatting issue.
- MINORtypoIntroduction, paragraph 5“Cigand Naziroglu”→ Cig Naziroglu or correct the author's nameThe citation appears to have a typo in the author's name.
- MINORclarityMaterials and methods, Inclusion and exclusion criteria“A total of two investigators developed the selection criteria”→ Two investigators developed the selection criteriaPhrase 'A total of' is redundant.
- MINORconsistencyTable II header“Case control study”→ Quality assessment of included studies using the Newcastle-Ottawa ScaleThe table header 'Case control study' is misleading as it includes cohort studies.
- MINORotherResults, Characteristics of the included studies“A total of four out of eight were cohort studies”→ Four of the eight studies were cohort studiesSlightly awkward phrasing.
- MINORotherAbstract, line 1“Shih Ya-Wen O'Brien Anthony Paul Hung Chin-Sheng Chen Kee-Hsin Hou Wen-Hsuan Tsai Hsiu-Ting”→ Insert commas between author names for clarity: 'Shih Ya-Wen, O'Brien Anthony Paul, Hung Chin-Sheng, Chen Kee-Hsin, Hou Wen-Hsuan, Tsai Hsiu-Ting'Author list is run together without punctuation.
- MINORgrammarIntroduction, paragraph 1“Exposure to ionizing radiation has been demonstrated to constitute a breast cancer risk”→ Change 'constitute' to 'increase': 'has been demonstrated to increase breast cancer risk'Unnatural phrasing.
- MINORconsistencyMaterials and methods, Data sources and search strategy“To increase the precision and specificity of article retrieval, [mesh term] and [text word] were used to search each databases.”→ Change 'each databases' to 'each database'Subject-verb agreement.
Published paper is methodologically robust but has a gap in reporting participant-level biological characteristics; readers should be aware that the meta-analysis does not provide a systematic breakdown of age, sex, or health status across studies. The missing p-value for the main effect is not a flaw given estimation-based reporting. The Newcastle-Ottawa Scale table labeling issue (Table II header says 'Case control study' but includes cohort studies) is a minor inconsistency that may warrant a correction.
- 1.HIGHrigorCorrect the mislabeling of Table II: the header currently reads 'Case control study' but includes cohort studies. Adapt the Newcastle-Ottawa Scale assessment to be appropriate for both study designs, or present separate tables. This is a methodological inconsistency that should be addressed in a correction.Using an identical quality assessment scale for case-control and cohort studies without adaptation may affect the validity of the quality assessment and could mislead readers.
- 2.HIGHreportingAdd a column to Table I explicitly reporting the sex of participants for each included study. If all studies are on women, state this clearly in the table.Sex is a fundamental biological variable; its absence from the characteristics table limits the ability to assess the applicability of the findings and potential confounding.
- 3.HIGHcopyeditCorrect the typo in the author name 'Cigand Naziroglu' in the Introduction (reference citation appears to have a misspelled author name).Incorrect author names in citations undermine the credibility of the reference list and could cause confusion for readers.
- 4.MEDIUMreportingAdd demographic and health status information (e.g., age range, race/ethnicity, comorbidities) to Table I if available from the original studies.Comprehensive reporting of participant characteristics enhances transparency and allows readers to evaluate the generalizability of the meta-analysis results.
- 5.MEDIUMcopyeditFix grammar errors: change 'each databases' to 'each database' in the Data sources and search strategy section; remove redundant 'A total of' phrases throughout the manuscript.These minor errors reduce the professional polish of the manuscript and could be distracting to reviewers and readers.
- 6.MEDIUMcopyeditAdd commas between author names in the abstract author list to improve clarity.The current run-together author names may cause confusion about individual names.
- 7.LOWreportingInclude a note in the Discussion about the absence of a formal search completeness assessment, acknowledging that the comprehensiveness of the search was not formally quantified.While the search strategy appears comprehensive, transparency about its limitations would strengthen the methodological disclosure.
- 8.LOWreportingIn the Introduction, elaborate on how the limitations of prior research (e.g., lack of dose-response data, exposure misclassification) are addressed in this meta-analysis.A more explicit discussion of how the current study improves upon previous work would strengthen the scientific premise.
- 9.LOWdata codeConsider adding a supplementary data file containing the extracted study-level data (e.g., effect sizes, confidence intervals, and study characteristics) to a repository for reproducibility.Although the data are reported in the article, providing a machine-readable table would facilitate independent re-analysis and meta-analytic updates.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.