# COSMOS prospective cohort of mobile phone users

> An international prospective cohort that records phone use from operator data rather than recall, designed to address the recall-bias problem that limits the case–control literature.

- Canonical URL: https://emfbase.com/studies/cosmos-cohort
- Collection: Research record
- Record id: `studies:cosmos-cohort`
- Status: provisional
- Last verified: 2026-08-12
- JSON: https://emfbase.com/api/studies/cosmos-cohort

## Key values

- **Authors:** COSMOS consortium
- **Year:** 2024
- **Design:** cohort
- **Exposure studied:** Operator-recorded call time and data use, with long-term follow-up for cancer and other outcomes
- **Place in the evidence base:** moderate
- **Primary document:** https://www.ukcosmos.org/

## Finding

Published analyses to date report no association between mobile phone call time and headache, tinnitus, sleep quality or, in reported cancer analyses, brain tumour incidence. Follow-up is ongoing.

## What it cannot establish

Follow-up is still short relative to plausible latency for brain tumours, and participants are volunteers. Consult the consortium's publication list for the current state rather than relying on any single result.

## Related records

- [WHO-commissioned systematic review of RF exposure and cancer in humans (2024)](https://emfbase.com/studies/who-rf-cancer-review-2024): A systematic review and meta-analysis of human observational studies found no association between mobile phone use and brain or head-and-neck cancers, including for long-term and heavy use.

## Sources

- `cosmos-2024` Imperial College London / COSMOS consortium. *COSMOS cohort study of mobile phone use and health outcomes*. https://www.ukcosmos.org/ — peer-reviewed, link checked 2026-08-12

Data licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Dataset version 2026.08.1. Retrieved from EMFBase, https://emfbase.com.

This record is reference data, not medical or legal advice. Regulatory limits, scientific guidelines and precautionary values are labelled separately and answer different questions.