"Exposure Equipment and Dosimetry

The design and dosimetry of the exposure equipment have been published previously (47, 48). A brief description follows.

Chamberettes. The RF-radiation exposure was done in chamberettes. The chamberettes were designed to expose 40 rats each. Each chamberette was equipped with an amplifier feeding four antennas that provided the electromagnetic signal and four carousels onto which rats in tubular restrainers were positioned. A rat restraint device consisted of a clear acrylic cylinder and a white Delrint retaining piece. An animal entered the restraint cylinder and the retainer piece was secured to the back of the restraint device. Ten restraint devices were then arranged symmetrically around the center antenna on a carousel. An air hose was attached to each carousel’s center to continuously provide fresh air from the irradiation laboratory. An exhaust fan mounted on the top of a chamberette maintained the internal environment within the chamberette. Lighting for each chamberette was provided by four 20 W fluorescent tubes, so that the normal light/dark cycle of the rats was not affected. The sham chamberettes were identical but without an RF power amplifier and input signal. The effects of restraint stress may modify a biological response, and efforts were made to minimize this stress as much as possible. Restraints were custom designed, and four different sizes of restraint devices were used to accommodate the animals as they grew naturally. In addition, rats were irradiated starting in the early morning hours to take advantage of their sleep cycle. The relative location of the antennas with respect to the animals was maintained for all restraint device sizes.

Antennas. Standard telecommunications antennas were used (Motorola model NAF5039A). They were specified to present over the 825–849 MHz mobile transmit frequency band, a constant 50-V impedance to the 50-V transmission line. The slot average power supplied to each antenna was between 1 and 2 W. Each antenna was driven by an RG142B/U type coaxial cable terminated in an SMA male plug; connection to the antenna was by way of an SMA female to SMA male right angle adapter.

CDMA source. The signal was O-QPSK modulated, centered around 847.74 MHz, and spread over a nominal bandwidth of 1.288 MHz. The unit used was a Qualcomm Beta 1 Prototype mobile phone (Type CD- 3000) with a frequency tolerance set to within 2.5 ppm. Generator output was spread to a four-way power divider, each output of which was then passed by way of a chain of discrete 0.5-dB attenuators and a coaxial cable to a 100 W microwave amplifier; the attenuator chains were used to balance the outputs of the four amplifiers. Each amplifier was then split four ways by another power divider, and the four outputs were fed to type N jacks on the side of a CDMA chamberette. Inside each chamberette, each signal was conveyed to its antenna by way of an RG142 coaxial cable.

FDMA source. The FDMA signal was a frequency-modulated continuous-wave transmission. The modulating signal was generated by a minidisk player. This signal modulated a frequency-synthesizing microwave generator (Hewlett Packard 8675A) whose output was spread by a fourway power divider. Each of the four divided signals was then passed by way of a chain of discrete 0.5-dB attenuators and a coaxial cable to a 50 W microwave amplifier. The attenuator chains were used to balance the outputs of the four amplifiers. The output of each amplifier was split four ways by another power divider. These four outputs were fed to type N jacks on the side of a chamberette. Inside each chamberette, each signal was conveyed to its antenna by way of an RG142 coaxial cable.

Dosimetry. Using the above equipment, thermographic and thermometric measures were performed and have been published previously

(48). The time-averaged specific absorption rate (SAR) in brain tissuemeasured thermometrically was 0.85 6 0.34 W/kg per watt of net (timeaveraged) input power into the radiating antenna; the time-averaged out put power for each antenna lay within the range 1.50 6 0.25 W. However, the actual spot SAR at a point within the brain was a function of the spot’s intracranial location, rat head position, and rat body weight. Over the course of a 2-year study, the inferred SAR could therefore vary from as little as 0.5 W/kg to as much as 2.5 W/kg, spanning the range of greatest interest for cellular telephones. In brief, a nominal time-averaged brain SAR value for this study was 1.3 6 0.5 W/kg (mean 6 SD)

Equipment stability assessment. On each day of irradiation, visual verifications that the chamberette lights, fan and air hoses were working were done and the results were recorded. After a chamberette was loaded with rats, the associated amplifier was turned on. It was verified visually and recorded that the signal sources and supplies were operating properly. The chamberette label indicated which amplifier was connected to the chamberette. The time the amplifier was turned on was recorded. After all of the chamberettes were loaded, the meters on the amplifiers were read and the maximum forward and reverse power readings (in dB) were recorded daily by the technical staff. These readings were summarized and the data monitored by the engineering team (co-authors Moros and Straube) on a regular basis to ensure stability. Additionally, inspections of the irradiation facility were performed by a senior engineer (Straube) before the start of the project, three times during the project, and once after the project ended. These inspections included verification of meter

readings, sentinel slot average power and antenna radiating verification.

Method of assignment of rats to treatment groups. Animals were first randomized so that body weight was evenly distributed among irradiation groups. Each irradiation group was permanently assigned to four chamberettes labeled A, B, C and D, each of which contained four shelves (labeled 1, 2, 3 and 4) capable of holding 10 rats that were oriented radially about a central dipole antenna. Each day, FDMA, CDMA and SHAM were loaded with the same sequence. However, a pseudo-random positioning algorithm was used to predetermine the day-by-day stochastic placement of each consecutively filled tray of 10 rats into a chamberette; five daily sequences were provided for 16 consecutive weeks, after which the sequence was repeated. This procedure was intended to randomize small

changes in the ambient environment based on the fact that it is not feasible to create a test design which enables 65% control over visible light, audible sound, mechanical vibration, air flow, and the variation from position to position of extremely low-frequency and direct-current fields, and specific absorption rates. If rats died naturally (or were euthanized due to illness) before the end of the experiment, the remaining rats continued to be loaded in sequence without gaps and any partially filled trays were completed with phantoms electromagnetically similar to the rats.

Necropsy

Animals were killed humanely with carbon dioxide and the body weights were measured. Body weight in rodents is one effective easure

of overall health and condition. Rats that are ill or under chronic stress will not maintain normal body weight. At necropsy all major organs were evaluated grossly and taken for histopathology including brain, spinal cord, adrenal glands, esophagus, heart, stomach, intestines, kidney, liver, lung, cervical lymph nodes, nasal cavity, reproductive organs, pancreas, salivary gland, spleen, thymus, thyroid, trachea, and urinary bladder. Any abnormalities in other organs were likewise noted and portions were taken for histopathological evaluation.

Tissue Fixation and Processing

Tissues were immersion-fixed in 10% neutral-buffered formalin. After fixation for at least 48 h, the tissues were embedded in paraffin and sectioned at 5 mm thickness. For brain tissue, every 80th section was mounted, and for the spinal cord three sections were mounted (cervical, lumbar and thoracic). For all other organs at least one section was mounted. All sections were stained with hematoxylin and eosin, covered with cover slips, and evaluated by the pathologist."

Fischer 344 rats, female

HAWC