In their just-posted comments on exposure issues in their study, the authors correct a trivial problem with units. But the study has far greater problems that were suggested by a previous correspondent. I have been communicating with the authors about these issues, so far without resolution.

1. The RF meter that the authors used for the study, the Tenmars TM-196, is an inexpensive consumer product that is demonstrably unacceptable for the measurements in the study. The meter was recently reviewed by a certified test lab (WILA, Bonn, Germany) . The report concluded ?Since the readings of this meter are all over the place, it is not even possible to obtain a rough estimate in the presence of higher exposure levels? the measurement characteristics are completely unacceptable across the entire frequency range tested.? (http://www.wilabonn.de/im...). Indeed, the authors report RF field levels near a cellular base station (a 100 ft. monopole with panel antennas for a single carrier at the top) that are two or more orders of magnitude above levels calculated using simple antenna theory for such a station. There are reports on the Internet by apparently qualified individuals about the tendency of the Tenmars meter to give vastly too high readings near cell base stations (https://www.youtube.com/w...).

Moreover, the RF exposure to the rats was determined by using the meter above its stated measurement range. The instruction manual for the meter lists its upper measurement range as 11 V/m, with an ?overload limit? of 17.7 V/m. The measurements reported in the paper were all above the maximum range, and in fact were close to the ?overload limit?.

The actual exposure to the rats was almost certainly higher. Simple calculations for the transmitter/antenna (a 1 W UHF RFID transceiver with a standard antenna with numeric gain of 4) show RF power densities several times higher than reported in the paper. I conclude that the RF exposure measurements in the paper are totally unreliable.

2. The authors report increases in ?skin temperature? of the irradiated rats of 2.1 C, measured with what they describe as a ?laser-thermometer (TW2, ThermoWorks,USA). In fact, that is an inexpensive consumer grade infrared temperature sensor that uses a laser for aiming. The device has 6:1 optics, which means that the measurement spot is 1 inch in diameter when the instrument is held at a distance of 6 inches from the target. It would take exceptional care to measure the surface temperature of a rat without having the measurements contaminated by environmental temperatures, and the authors do not state what precautions they took. At best the instrument would measure the temperature of the rat?s *fur*, not skin. The website of a company that makes an infrared temperature monitor for rodents warns ? "Important notice: measurement taken in the animal fur may include the ambient air caught in the hair, thus creating important measurement variations. It's therefore advisable to measure on naked surfaces, like skin, tail, nose, genital area...? (http://www.braintreesci.c...), which was not done in this study. A previous correspondent noted that the RF exposure was far too low to produce the 2.1 C increase in skin temperature the authors claim. However, hair is lossy at UHF frequencies and it may have been heated by the comparatively low exposures without a similar increase in skin temperature.

These are serious issues that compromise the validity of the study, to the point that its results are essentially uninterpretable.

Kenneth R. Foster
Professor of Bioengineering
University of Pennsylvania