Headache in the international cohort study of mobile phone use and health (COSMOS) in the

Abstract

Headache in the international cohort study of mobile phone use and health (COSMOS) in the Netherlands and the United Kingdom My note: This cohort study suffers from numerous methodological problems that undermine the authors' results and conclusions. Traini E, Smith RB, Vermeulen R, Kromhout H, Schüz J, Feychting M, Auvinen A, Poulsen AH, Deltour I, Muller DC, Heller J, Tettamanti G, Elliott P, Huss A, Toledano MB. Headache in the international cohort study of mobile phone use and health (COSMOS) in the Netherlands and the United Kingdom. Environmental Research. 2024. doi: 10.1016/j.envres.2024.118290. Highlights • Prospective study exploring the relationship between mobile phone use and headache. • Associations with call-time were largely explained by texting. • Associations likely to reflect behavioural aspects of mobile phone use. Abstract Headache is a common condition with a substantial burden of disease worldwide. Concerns have been raised over the potential impact of long-term mobile phone use on headache due to radiofrequency electromagnetic fields (RF-EMFs). We explored prospectively the association between mobile phone use at baseline (2009–2012) and headache at follow-up (2015–2018) by analysing pooled data consisting of the Dutch and UK cohorts of the Cohort Study of Mobile Phone Use and Health (COSMOS) (N = 78,437). Frequency of headache, migraine, and information on mobile phone use, including use of hands-free devices and frequency of texting, were self-reported. We collected objective operator data to obtain regression calibrated estimates of voice call duration. In the model mutually adjusted for call-time and text messaging, participants in the high category of call-time showed an adjusted odds ratio (OR) of 1.04 (95 % CI: 0.94–1.15), with no clear trend of reporting headache with increasing call-time. However, we found an increased risk of weekly headache (OR = 1.40, 95 % CI: 1.25–1.56) in the high category of text messaging, with a clear increase in reporting headache with increasing texting. Due to the negligible exposure to RF-EMFs from texting, our results suggest that mechanisms other than RF-EMFs are responsible for the increased risk of headache that we found among mobile phone users. Excerpts A study conducted in Sweden and Finland as part of COSMOS found limited evidence for an association between weekly headache and the highest level of mobile phone use and no clear trend with increasing call-time (Auvinen et al., 2019). The association of headache with call-time appeared stronger for calls via the Universal Mobile Telecommunication System (UMTS) (3G) network than via the older Global System for Mobile (GSM) (2G) telecommunications technology, despite the latter involving higher RF- EMF exposure levels to the head (van Wel et al., 2021) .... Headache has been linked to excessive mobile phone use, but the mechanism by which mobile phone use may cause symptoms is not properly understood (Wang et al., 2017; Cerutti et al., 2016; Frey, 1998; Hocking, 1998; Oftedal et al., 2000; Schoeni et al., 2015). Previous research in adolescents has suggested that other exposures related to mobile phone use, but not exposure to RF-EMFs, should be considered the causal factor for various symptoms, as the strongest associations were found with activities that cause minimal RF-EMF exposure to the head, such as texting or gaming (Schoeni et al., 2017). Other studies have indicated that stress or unfavourable usage, such as late-night use, may be associated with an increase in reported health symptoms, such as headache (Szyjkowska et al., 2014; Röösli, 2008; Thomée et al., 2011). It is therefore crucial to distinguish between using a mobile phone for calling and other activities that expose the brain to RF-EMFs at lower levels, such as Internet browsing (Cabré-Riera et al., 2022a; SSM's Scientific Council on Electromagnetic Fields, 2020). Our study attempted to disentangle the exposure-outcome gradient by considering call-time as a proxy for RF-EMF exposure and texting as a proxy for usage with negligible RF-EMF exposure to the brain (Wall et al., 2019). This study's mobile phone usage data was gathered between 2009 and 2012. During those years, texting was the most popular activity unrelated to RF-EMF exposure..... The Headache Impact Test (HIT-6) score with a cut-off of 56 points defined severe weekly headache. The HIT-6 is a tool used to measure the impact headaches have on one's ability to function in various aspects of daily life, including work, school, home, and social contexts. The score, ranging from 36 to 78 points, provides a measure of the degree to which headaches affect daily life and functioning, with higher scores indicating a more significant impact on the participant's overall life (Kosinski et al., 2003).... Results from two-exposure models mutually adjusting for both call-time and texting at baseline, showed substantially lower risk estimates for weekly headache in the high call-time (RC-hfa) category (OR = 1.04, 95 % CI: 0.94–1.15), and no evidence of a trend (P trend = 0.292) (Table 2). Associations with texting were robust to adjustment for call-time: we observed an increased risk of weekly headache in the high category of texting (OR = 1.40, 95 % CI: 1.25–1.56) and a trend of increasing risk with increasing texting frequency (P trend<0.001), in line with results from the single-exposure model (Table 3).... In this large international prospective cohort of mobile phone users in the Netherlands and the UK, mobile phone use for calling and texting at baseline was associated with headaches at follow-up. Mutually adjusting for both call-time and texting considerably attenuated risk estimates for call-time, while associations with texting were still strong and robust to adjustment, with a clear exposure- outcome gradient. Headache has been linked to excessive mobile phone use, but the mechanism by which mobile phone use may cause symptoms is not properly understood (Wang et al., 2017; Cerutti et al., 2016; Frey, 1998; Hocking, 1998; Oftedal et al., 2000; Schoeni et al., 2015). Previous research in adolescents has suggested that other exposures related to mobile phone use, but not exposure to RF-EMFs, should be considered the causal factor for various symptoms, as the strongest associations were found with activities that cause minimal RF-EMF exposure to the head, such as texting or gaming (Schoeni et al., 2017).... Our study attempted to disentangle the exposure-outcome gradient by considering call-time as a proxy for RF-EMF exposure and texting as a proxy for usage with negligible RF-EMF exposure to the brain (Wall et al., 2019). This study's mobile phone usage data was gathered between 2009 and 2012. During those years, texting was the most popular activity unrelated to RF-EMF exposure. In both scenarios, we found an increased risk of headache in the high exposure category of mobile phone use with a positive exposure-outcome gradient confirmed by the test for trend. The attenuated risk estimates for call-time in the mutually adjusted model argue against an effect of exposure to RF- EMFs due to the negligible exposure attributed to texting. This conclusion is also supported by comparing call-time analyses with and without hands-free adjustment, where no risk reduction was found among users for the adjusted exposure metrics.... Regarding secondary health outcomes, we found consistent patterns of results for severe weekly headache and migraine at follow-up in terms of increased risk estimates and significant trends. Increasing risk of daily headache was associated with increasing texting (P trend<0.001) but not with increasing call-time (P trend = 0.448) (Table 2, Table 3). In both scenarios, we found an increased risk of headache in the high exposure category of mobile phone use with a positive exposure-outcome gradient confirmed by the test for trend. The attenuated risk estimates for call-time in the mutually adjusted model argue against an effect of exposure to RF- EMFs due to the negligible exposure attributed to texting. This conclusion is also supported by comparing call-time analyses with and without hands-free adjustment, where no risk reduction was found among users for the adjusted exposure metrics. In this study, the distribution of the exposure, specifically regarding call-time, differed between Dutch and UK participants. Mobile phone usage behaviour across countries cannot be assumed to be identical due to various factors such as cultural, economic, technological, and market dynamics (Böhm, 2015). To assess the consistency of our findings, we showed that defining the top exposure category for call-time based on the 80th percentile cut-off, thereby ensuring the inclusion of Dutch participants in the “high” exposure category, yielded results consistent with those obtained using the 90th percentile as a cut-off. These findings suggested that the association we found between call-time and headache was driven not only by UK but also Dutch participants. Of note, all analyses were adjusted for country of residence.... Our study also has limitations. First, we did not have information about “true” RF-EMF exposure. Exposure to RF-EMFs emitted by wireless devices is difficult to quantify, particularly in large populations and over extended periods, as it depends on different factors, such as reception quality or other factors influencing signal strength. In our study, we calculated several exposure metrics as proxies for RF-EMF exposure, which allowed us to estimate the average individual RF-EMF exposure in the population. Additionally, information on other aspects of usage, such as screen time, blue light exposure or unfavourable use at night, may be helpful to include in future studies. For highly transient and acute symptoms such as headache, using the peak of RF-EMF exposure might be theoretically preferable over the weekly exposure assessed in our study. However, adopting this approach would require substantially different exposure assessment methods that are impractical for large cohort studies, such as asking participants to regularly fill in a detailed usage diary. Given the study design and methodology used to assess RF-EMF exposure in COSMOS, the analysis of the association between RF-EMF peak exposure and reporting of headache symptoms was precluded. In light of the transient nature of headaches, future research may explore the potential effect of peak RF-EMF exposure on symptom onset more thoroughly. The composition of the Dutch cohort is not representative of the adult population of the Netherlands with respect to sex and age.... Conclusions In summary, we found that the use of mobile phones, particularly texting, is associated with headaches and migraines, and the associations with call-time were largely explained by texting. As the associations are driven more by text messaging than call-time, they do not appear to be explained by RF-EMF exposure from the mobile device but are likely to reflect lifestyle, other exposures, or behavioural factors associated with mobile phone use. Declaration of competing interest [Note that text is missing from this paragraph.] The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Maria Feychting reports financial support was provided by Swedish Research Council. Maria Feychting reports financial support was provided by AFA Insurance. Maria Feychting reports financial support was provided by Swedish Research Council for Health, Working Life and Welfare. Maria Feychting reports financial support was provided by Swedish Radiation Safety Authority. Maria Feychting reports financial support was provided by Sweden's Innovation Agency. Paul Elliott reports financial support was provided by Mobile Telecommunications and Health Research. Paul Elliott reports financial support was provided by National Institute for Health and Care Research Health Protection Research Unit. Anssi Auvinen reports financial support was provided by National Technology Agency (TEKES). Anssi Auvinen reports financial support was provided by Pirkanmaa Hospital District. Anssi Auvinen reports financial support was provided by Yrjö Jahnsson Foundation. Anssi Auvinen reports financial support was provided by Mobile Manufacturers Forum. Roel Vermeulen reports financial support was provided by Netherlands Organisation for Health Research and Development. Aslak Harbo Poulsen reports financial support was provided by the Danish Strategic Research Council. Joachim Schuz reports financial support was provided by French Agency for Food, Environmental and Occupational Health & Safety (ANSES). Maria Feychting reports a relationship with World Health Organization that includes:. Maria Feychting reports a relationship with Public Health England Advisory Group on Non-ionizing Radiation that includes:. Maria Feychting reports a relationship with Norwegian Institute of Public Health that includes:. Maria Feychting reports a relationship with Forskningsrådet för hälsa arbetsliv och välfärd that includes:. Maria Feychting reports a relationship with Swedish Radiation Safety Authority that includes:. Maria Feychting reports a relationship with International Commission on Non-Ionizing Radiation Protection that includes:. Aslak Harbo Poulsen reports a relationship with Swedish Radiation Safety Authority that includes:. Anssi Auvinen reports a relationship with International Commission on Non-Ionizing Radiation Protection that includes:. Anssi Auvinen reports a relationship with Scientific Committee on Emerging and Newly Identified Health Risks that includes:. Anssi Auvinen reports a relationship with European Commission, Swedish Radiation Safety Authority that includes:. Anssi Auvinen reports a relationship with World Health Organization that includes:. Anke Huss reports a relationship with Swedish Radiation Safety Authority that includes:. Anke Huss reports a relationship with International Commission on Non-Ionizing Radiation Protection that includes:. Anke Huss reports a relationship with BioEM society that includes:. Anke Huss reports a relationship with Gezondheidsraad that includes:. Anke Huss reports a relationship with Swiss Research Foundation for Electricity and Mobile Communication that includes:. Hans Kromhout reports a relationship with Gezondheidsraad that includes:. Hans Kromhout reports a relationship with World Health Organization that includes:. Mireille B Toledano reports a relationship with UK Committee on Medical Aspects of Radiation in the Environment (COMARE) that includes:. Open access paper: sciencedirect.com

AI evidence extraction

Main findings

In pooled Dutch and UK COSMOS cohort data (N=78,437), mutually adjusted models showed no clear association between higher voice call-time (regression-calibrated operator data) and headache at follow-up (high call-time OR 1.04, 95% CI 0.94–1.15; no clear trend). Higher texting frequency (self-reported) was associated with increased risk of weekly headache (high texting OR 1.40, 95% CI 1.25–1.56) with a positive trend; similar patterns were reported for severe weekly headache and migraine, and daily headache showed a trend with texting but not call-time.

Outcomes measured

• Headache frequency (including weekly headache)
• Severe weekly headache (HIT-6 cutoff 56)
• Migraine
• Daily headache

Limitations

• Headache, migraine, and mobile phone use variables (including texting and hands-free use) were self-reported (except operator data used for regression-calibrated call duration).
• No information on 'true' RF-EMF exposure; exposure metrics were proxies and RF-EMF exposure is difficult to quantify (e.g., depends on reception quality/signal strength).
• Weekly exposure metrics were used; peak/acute RF-EMF exposure relevant to transient symptoms like headache was not assessed and was considered impractical in this cohort design.
• Lack of information on other usage-related factors (e.g., screen time, blue light exposure, late-night use) noted as potentially important for future studies.
• Dutch cohort not representative of the adult Netherlands population with respect to sex and age.
• Mobile phone usage behavior differed between countries; analyses adjusted for country and sensitivity analyses with different cutoffs were described.

Suggested hubs

• who-icnirp (0.32)
  Competing interest statement notes relationships with WHO and ICNIRP among authors; study concerns RF-EMF health outcomes.

{
    "study_type": "cohort",
    "exposure": {
        "band": "RF",
        "source": "mobile phone",
        "frequency_mhz": null,
        "sar_wkg": null,
        "duration": "Baseline mobile phone use assessed 2009–2012; follow-up headache assessed 2015–2018"
    },
    "population": "Participants in the Dutch and UK cohorts of the COSMOS study (mobile phone users)",
    "sample_size": 78437,
    "outcomes": [
        "Headache frequency (including weekly headache)",
        "Severe weekly headache (HIT-6 cutoff 56)",
        "Migraine",
        "Daily headache"
    ],
    "main_findings": "In pooled Dutch and UK COSMOS cohort data (N=78,437), mutually adjusted models showed no clear association between higher voice call-time (regression-calibrated operator data) and headache at follow-up (high call-time OR 1.04, 95% CI 0.94–1.15; no clear trend). Higher texting frequency (self-reported) was associated with increased risk of weekly headache (high texting OR 1.40, 95% CI 1.25–1.56) with a positive trend; similar patterns were reported for severe weekly headache and migraine, and daily headache showed a trend with texting but not call-time.",
    "effect_direction": "mixed",
    "limitations": [
        "Headache, migraine, and mobile phone use variables (including texting and hands-free use) were self-reported (except operator data used for regression-calibrated call duration).",
        "No information on 'true' RF-EMF exposure; exposure metrics were proxies and RF-EMF exposure is difficult to quantify (e.g., depends on reception quality/signal strength).",
        "Weekly exposure metrics were used; peak/acute RF-EMF exposure relevant to transient symptoms like headache was not assessed and was considered impractical in this cohort design.",
        "Lack of information on other usage-related factors (e.g., screen time, blue light exposure, late-night use) noted as potentially important for future studies.",
        "Dutch cohort not representative of the adult Netherlands population with respect to sex and age.",
        "Mobile phone usage behavior differed between countries; analyses adjusted for country and sensitivity analyses with different cutoffs were described."
    ],
    "evidence_strength": "moderate",
    "confidence": 0.7800000000000000266453525910037569701671600341796875,
    "peer_reviewed_likely": "yes",
    "keywords": [
        "COSMOS",
        "mobile phone use",
        "RF-EMF",
        "radiofrequency electromagnetic fields",
        "call-time",
        "text messaging",
        "texting",
        "headache",
        "migraine",
        "HIT-6",
        "prospective cohort",
        "Netherlands",
        "United Kingdom",
        "operator data",
        "regression calibration"
    ],
    "suggested_hubs": [
        {
            "slug": "who-icnirp",
            "weight": 0.320000000000000006661338147750939242541790008544921875,
            "reason": "Competing interest statement notes relationships with WHO and ICNIRP among authors; study concerns RF-EMF health outcomes."
        }
    ]
}

Comments

Log in to comment.

No comments yet.