IEEE ICES Database
ElectroMagnetic Field Literature
Search Engine
IEEE ICES website

EMF Study
(Database last updated on Sep 25, 2022)

ID Number 1082
Study Type Human / Provocation
Model 900, 1800 MHz (GSM), 2100 MHz (W-CDMA) exposure to human subjects and analysis of subjective complaints and cognitive function

Human subjects were exposed in an anechoic chamber to 945 MHz (900GSM), 1840 MHz (1800GSM), or 2140 MHz (2100UMTS) with the antenna located 3 meters behind the subject (to simulate a base station-type exposure) producing an incident field strength of ~1 V/m - corresponding to peak SARs in the head ranging from 0.000045 to 0.000082 W/kg (over 10 grams) and 0.316 to 0.513 nW/kg (total head average). There were two groups of 36 subjects each tested under double-blind randomized conditions. Group A (hypersensitive) consisted of subjects who had registered complaints regarding GSM hypersensitivity with the Monitoring Network for Environmental Health in The Netherlands; Group B (reference group) consisted of subjects recruited from within TNO and via newspaper ad. Cognitive function (using a Taskomat protocol) and subjective complaints (using a questionnaire format) were determined. The authors reported significant increases in certain subjective complaints with exposure, most notably with 2100 UMTS, but these were not consistent between groups A and B. In addition, the authors reported an increase in reaction time (decrease in performance) with 900GSM in group A and a similar increase in reaction time with 2100UMTS in group B. A decrease in general reaction time (dual tracking) in group B was found with 1800GSM. A decrease in memory function in group B with 1800GSM and 2100UMTS, as well as a decrease in filtering ability with 900GSM exposure in group A was also reported.

Findings Effects
Status Completed Without Publication
Principal Investigator Ministry of Health & Welfare, The Netherlands
Funding Agency TNO, The Netherlands
  • Zwamborn, et al. TNO Report (FEL-03-C148), (2003) :-
  • Comments

    The subject pool may have been biased, as half had registered GSM complaints with the Monitoring Network for Environmental Health in The Netherlands and the other half responded to a messages that seems to have directly stated the nature of the study. Subjective complaints and complaint history were assessed by questionnaire, but it is not clear that it was blinded either, leading to further bias. The results showing an increase in reaction time (decrease in performance) are in direct contrast with Preece et al (who reported an increase in performance), and were small in magnitude (on the order of 3-4% change vs. control conditions). In addition, there was no consistency in the effect on reaction time between group A (effect with 900GSM) and group B (effect wth 2100UMTS). For subjective complaints there was also no consistency in the data between groups A and B (see Table 12.1 in conclusion section). The authors suggested previous negative reports by Hietanen et al may have been flawed by not having a sufficiently low ambient RF environment (below 2 mW/cm2), although the effect of the computer screen in the TNO study did not seem to be accounted for in the description of ambient fields for the test chamber. For the cognitive function tests, there was a unique battery of 30 different tests that cannot be directly compared to results of Preece et al and Koivisto et al. Since only one exposure level was used, no "dose response" data are available. In a letter dated October 7, 2003, Dr. Zwamborn stated his disassociation with inaccurate suggestions in the media that their results show that UMTS-like signals cause significant harmful effects to health or that these fields cause headaches or other health related issues. He also stated the necessity for his results to be independently reproduced before any justified conclusions can be drawn. A review by the the Health Council of the Netherlands in June 2004 ( subsequently concluded that TNO report was of good quality in terms of design and execution, but The Committee had some concerns, however, regarding the questionnaire, interpretation and statistical evaluation of the data, and whether the exposure was relevant to normal real life exposures. The review recommended replication.