INTRODUCTION.

In the last few decades, the use of devices which emit electromagnetic fields (EMFs) has increased considerably. This proliferation has been accompanied by an increased concern about possible health effects of exposure to these fields (for reviews see 1, 2, 3, 4). Human-made sources in the spectrum of EMFs (up to 300 000 MHz) produce local field levels many orders of magnitude above the natural background. Radio and television transmitters are examples of radiofrequency (RF - 0.1 - 300 MHz) and microwave (MW - 300 - 300 000 MHz) sources that intentionally produce EM emmisions to the environment. Exposure levels in the MW range are traditionally described in terms of "power density" and are normally reported in watt per square metre (W/m²), or milliwatts per square centimetre (mW/cm² = 10 W/m²). Systems which emit MW energy to the environment, including broadcasting stations, radar and telecommunication antennas, vary greatly in terms of their design. This diversity results in somewhat different approaches in evaluating human exposure and potential risk problems.

Fortunately, a comprehensive evaluation of residential exposure to RF and MW indicates that, in general, the exposure levels are relatively low. Measurements performed in 15 large cities in the USA revealed that the median exposure level ranged about 0.05 W/m², with 90% of residents being exposed to fields not exceeding 0.1 W/m² (4). Only approximatly 1% of the population studied was potentially exposed to levels greater than 0.1 W/m² . These higher exposures occur at limited areas located close to strong MW sources. Such situations can exist e.g. in proximity to very powerful, ground-level transmitters. or to low-power, in-town repeaters, which are typically mounted on the top of tall buildings (1, 5).

Introduction of cellular phone (CP) systems and a fast increase of number of users of hand-held phones in the last decade has changed the MF exposure levels of the population quite considerably. With CPs, a MW transmitter has been for the first time ever in history put right up against the side of anyone’s head, and switched on. Analysis of distribution and absorption of the radiation revealed that about 40% of the MW energy emitted from CP antenna goes into the user’s head and hands (6). Such situation raised immediately concerns about possible health risks of the exposures, both among the bioelectromagnetic community and the public.

Fortunately, till now, no cases of specific syndromes and/or increased risk of any spontaneous diseases, which can be causally linked to action of MWs emitted from CPs, have been documented in medical or epidemiological observations of CP users (6, 7, 8, 9). On the other side, a variety of non-specific health symptoms (NSHS), including headaches, fatigue, affected sleep parameters, and small changes in blood pressure, generally referred as „MW hypersensitivity” were reported in single studies of CP users (10, 11, 12, 13). Still more, the recent epidemiologic study performed in Sweden and Norway on about 11 000 CP users points a clear relation between occurance of NSHS and intensity of CP use – about 20% of these who use CP for 1 hr daily or more report NSHS, while the numbers are considerably lower for those who use CP for 15 or 30 min daily (14). From the scientists’s point of view, the present bulk of evidence allows to conclude without doubts that CPs have a „biological effect”, although the clinical relevance of this „effect” still remains unclear. On the other side, the CP industry has conducted a sophisticated and, so far, very successful campaign to accentuate the positive possibility that their product do not pose any health problems. The problem of health risks and mobile phone systems has been discussed in detail in a comprehensive and well balanced report of the Independent Expert Group on Mobile Phones, headed by Sir William Stewart (3). The report has been published in 2000 by NRPB in UK and is addtionally available at the website address www.iegmp.org.uk.

MW energy penetrates inside biological tissues and may be there transformed into heat. The deposition of larger amounts of MW energy in the human body tends to increase the body temperature to the extent which depends on numerous external and internal factors, including thermal environment and metabolic heat production. In normal thermal environment a SAR (specific absorption rate) of MW radiation of 1 - 4 W/kg for 30 minutes produces an average body temperaurte increase of less than 1⁰C for healthy adults (2, 4) a situation which is considered as tolerable from the thermophysiologic point of view. Thermal effects caused by absorption of MW energy are considered as dangerous (2) and therefore, all safety guidelines leave a large margin of protection against such complications, even at the worst possible environmental conditions.

A real problem starts when the effects of low-level MW fields are being considered, when the amount of absorbed energy is much too small to cause a detectable increase of temperature (15). At present there exist sufficient experimental and epidemiological evidence which clearly indicates that under certain conditions of exposure, weak and very weak MW fields can cause measurable effects in biological organisms (cells, animals, human beings) (for most recent reviews, see 3, 6, 7, 15, 16), but mechanisms of these effects and their relevance for health status of the organisms are still difficult for interpretation. In general, bioeffects linked with eposures in low-level MW fields are weak, transient, disappear shortly after exposure and are difficult for replication. Still more, quite frequently certain bioeffects (e.g. increase in efflux of calcium ions from brain tissues and neural cells exposed to weak MW fields modulated at 16 Hz frequency (17) occur only at particular frequencies and/or modulations of MW ("windowing" phenomena) and/or do not show any relation with exposure intensity (dose).

Analysis of available publications on effects of exposure in low-level MW fields reveals that only part of the material fulfills the quality assurance criteria (15). Numerous reports from experimental studies do not provide sufficient information about exposure conditions, MW dosimetry, use of sham-exposed controls and/or experimental design. In numerous medical and epidemiological observations of personnel working in MW fields, the assessment of individual exposure is at least uncertain, particularly historical exposures are often determined via surrogates (e.g. job titles, service cards, etc.) (18). In view of the above, a considerable practical experience in studies of bioeffects of EMFs is needed for critical evaluation of the available literature on effects of low-level MW fields, synthesis of data and selection of results into these which are properly verified, possible but requiring further confirmation from those which are still poorly documented and seem unacceptable for experienced reviewers (Table I).

Non-cancer health disorders which are frequently linked to long-term exposures in low-level MW fields include a variety of behavioural effects, functional changes in the neural and circulatory systems, dysregulation of the autonomic control of internal organs, as well as certain ocular effects (for recent reviews, see 7, 8, 11, 16). Although there exists partial support for these effects from experimental studies in animals and medical observations of occupational groups which are exposed to MW at work, it still remains an open question whether or not such health disorders remain causally linked to the radiation or are the effect of non-specific stress reactions experienced in the work environment. Another concept which may explain the above health disorders occuring after exposure in weak MW fields is generally referred as "EMF hypersensitivity", which is defined as ability of individual people to react to EMFs at significantly lower levels than normal (13). According to its etiology, in general the term "EMF hypersensitivity" (EMH) is used for people who claim to have subjective health problems (headaches, depressive symptoms, neurasthenia, anxiety, etc.) due to the nearby electric and electronic appliances and/or MW transmitters. The problem of EMH has become recently a subject of systemic investigations, as in some European countries (mostly in the Nordic countries and in Germany) centers of occupational medicine have to deal with large number of persons claiming to be hypersensitive and several self-aid groups of citizens were already formed (13, 15). Existing studies, based on response of "hypersensitive" subjects to provocational exposures in weak EMFs, tend to support the existence of EMH, but only in part of those who claim their hypersensitivity. EMH is assumed to occur, with more or less pronounced symptoms, in about 1-2% of the general population, but in most

Table I.

Biological effects and health hazards of long-term exposures

in low-level microwave fields.

Experimental investigations

Human studies

A. Effects with sufficient confirmation in experimental/human studies,

occuring frequently under different conditions of exposure.

· No documented health effects and delayed risks of long-term exposure in low-level MW fields;

· Generally weak and poorly replicable effects in experimental animals exposed in low-level MW fields;

· Inability to establish thresholds for bioeffects related to exposure in low-level MW fields;

· Non-specific symptoms without proven clinical relevance;

· Dysregulation of autonomic control of physiologic systems;

· Functional changes in brain bioelectric activity without clinical relevance.

B. Probable effects, difficult for confirmation and/or replication,

occuring under specific conditions of exposure („widowing” phenomena).

· Influenced (generally slightly increased) efflux of calcium from cells and tissues exposed in vitro;

· Slightly changed active membrane transport of ions in cells exposed in vitro;

· Activation of certain cellular enzymes, relevant for proliferation and/or carcinogenesis (e.g. ornithone decarboxylase, protein kinases);

· Changed reactivity of immunocompetent cells (lymphocytes, killer cells, macrophages) after exposure in vitro;

· Individual hypersensitivity to EM fields;

· Non-specific stress reaction with general adaptation syndrome (GAS);

· Intensification of symptoms of liability of autonomic regulation of physiologic systems („vegetative neuroses”).

C. Poorly documented effects, frequently not replicable,

relation to exposure conditions needs confirmation.

· Increased neoplastic transformation rate of cells exposed in vitro;

· Faster development of spontaneous and/or induced neoplasms in mice and rats;

· More pronounced cytopathic effects after exposure in pulse-modulated MW fields than in equivalent continous wave fields.

· Increased risk of development of certain neoplastic diseases (haemopoietic and lymphatic malignancies, brain tumours);

· Increased risk of development of organic diseases, influenced by neurogenic and psychic stimuli;

· Immunologic and endocrine abnormalities.

D. Unconfirmed/unacceptable effects reported in the literature,

difficult for interpretation and/or validation.

· Morphologic injury of cells and/or internal organs;

· Specific diseases (e.g. „microwave” sickness)

cases the symptoms can be triggered by exposure in power frequency (50 Hz) EMFs, while MW seem to play only little role in development of the symptoms (8, 13). It should be however emphasized that current data are still unable to give a definite positive identification of causal relationships between exposure in weak EMFs and appearance of non-specific health symptoms, even in subjects concerned as "hypersensitive". Improved methodologic design of future studies with healthy volunteers and "hypersensitive" subjects and better coordination of studies in different countries may allow to solve the problem.

CARCINOGENIC POTENCY OF MICROWAVE RADIATION.

Electromagnetic fields have been linked with increased risk of neoplastic diseases for a long time, but the available experimental and medical data still did not allow for valid conclusions. There exists a fragmentaric and scarce support from experimental studies which indicates a possibility of epigenetic (non-genotoxic) potency of microwave energy in the multistep process of carcinogenesis, although possible mechanisms underlying these phenomena still remain hypothetic. A detail analysis of this problem is presented in the recent IEGMP-2000 Report (3).

Human data on possible health effects of exposure to low-level MW fields are mostly based on medical and epidemiological observations of occupational groups which are exposed to MW at work (military personnel, police officers, physiotherapists who use medical diathermy equipment, plastic sealers and workers in the broadcasting, transport and communication centeres). Effects of residential exposures were relatively rarely a subject of sudy. The epidemiological studies completed so far have mostly looked at cancer incidence in residents living close to radio and television transmitters and did not found a sufficient evidence for an increased risk. Following a study of residents living around one TV and radio broadcasting tower in UK in which a significant increase in morbidity from adult leukaemia was reported in people residing within 2 km of the transmitter (19), a more comprehensive study, performed by the same authors around 20 transmission towers in UK, did not confirm this finding (20). The study, based on 79 cases of adult leukemia revealed that for persons residing within 2 km from the transmitters the morbidity ratio was not increased (ebserved/expected O/E = 0.97), however a small, but significant, decline in risk of adult leukemia with distance from transmitters in the 2-10 km. was found (19, 20). Similar observations were made in Australia. A study of cancer incidence among residents living in the "inner" (close to TV towers) and "outer" (more distant) municipalities in Northern Sydney reported an increased morbidity and mortality of childhood leukaemia (21) in the "inner" municipalities. However, when these data were reanalyzed and other "inner" municipalities were added (22), it apperared that the excess of childhood leukaemia was restricted only to one (of six) "inner" municipalities and there exist no evideneces for linking it with the low-level MW exposures.

Epidemiological observations of occupational groups which are exposed to MW at work also do not provide sufficient evidence for a causal links between exposure and increased risk of neoplastic diseases, although in some studies a considerably higher morbidity rates were reported (for recent reviews, see 3, 23). It should be also pointed that each work environment has an individual combination of physical, chemical and psychosocial factors which may influence human physiology, including development of neoplastic diseases, in a very specific and unique way. Therefore, the results of occupational studies of MW-exposed workers cannot be directly extrapolated as health risks for the general public, the more that intensities and time sequences of MW exposures in workers and in the environment are different (18, 24). A typical MW intensities at work range from 2 - 10 W/m² with incidental exposures at 10 - 30 W/m² and a period of exposure being limited to 1-2 hr during a working shift (4), while in the environment and homes MW fields normally do not exceed 0.1 W/m², but the exposure tends to be continous.

EPIDEMIOLOGICAL STUDIES ON POLISH MILITARY PERSONNEL.

Some time ago the results of our retrospective analysis of cancer morbidity for the whole population of career military personnel in Poland during the decade of 1970 - 1979 was published (24), although at that time the exact size of the population could not be revealed; therefore, the results and their discussion were limited to mortality rates (number of newly diagnosed cancer cases per 100 000 of subjects per year). Nevertheless, a significantly higher rate of particular types of neoplasms (haematologic, lymphatic system, skin tumours, alimentary tract cancers) in personnel exposed occupartionally to RFs and MWs (24) encouraged us to continue the prospective analysis of morbidity and extend the observation period for the years 1980 - 1985. In 1996 the joint analysis covering the 15- year period of 1971 - 1985 has been published (25). It has been found that the subpopulation of about 3-4% which had a documented occupational exposure to RF/MW radiation developed about 9% of all malignancies, giving the OER (Observed/Exposed Ratio) of 2.1 - 3.1, depending on year of analysis. This difference in cancer morbidity related only to particular types of malignancies and still more, the retrospective analysis did not allow for precise assessment of past RF/MW exposure intensity (dosis). Therefore, at that time the search for possible relations between cancer morbity (risks) and levels of the RF/MW exposure was not possible. Additionally, we were aware that the analysis was based on generally low number of registered cases of neoplasms and both increasing size of the RF/MW-exposed population and longer period of observation has been postulated, before final conclusions can be obtained.

In 1985 a prospective analysis of cancer morbidity in Polish military career personnel has been started and additionally, the exposure levels of the personnel were measured.

In this paper we present the available data and the summary of cancer morbidity for the 20-year period (1971 - 1990), including the 5-year period (1986 - 1990) where both prospective analysis of new cases and assessment of the RF/MW exposure were performed.

Material and Methods.

During 1971 - 1985 a retrospective analysis of all cases of neoplasms in the whole population of career personnel in Polish army have been collected and compared with service records listing exposure to RF/MW. Starting from 1985 each newly diagnosed case of malignancy is recorded and a questionnaire describing type of service and possible exposure to different risk factors, including RF/MW radiation is filled. Additionally, all service posts where exposure of personnel to RF/MWs occurs were assessed for the exposure levels. This allowed to divide the RF/MW-exposed posts into four types - low (1 - 2 W/m²), medium (2 - 6 W/m²), above medium (6 - 10 W/m²) and high (exceeding 10 W/m²) maximal exposure levels during working shift (Table V).

The size of population analyzed ( active servicemen aged 25 - 59 years) varied from year to year with a mean of 124 500 and standard deviation of 18 300. About 3% of this population (3 500 - 4 500 subjects) were considered on base of service records as exposed to RF/MW radiation. For analysis the population was divided into four age groups (< 30, 30 - 39, 40 - 49, 50 - 59).

All cases of neoplasms diagnosed in the population were divided into 12 groups (type/localization - Table III). Additionally, analysis of morbidity was performed in age groups and for the prospective study (1986 - 1990) also in subgroups with various levels of RF/MW exposure.

A battery of statistical tests was used for analysis and correlation of the results. A CSS Statistica 5.5. (Statsoft) software was applied for computing the results.

Results and Discussion.

. During the 20-year period (1971 - 1990) in the investigated population a total of 2 493 various neoplasms was diagnosed, 2 355 (94.46%) of them in non-exposed subjects and 138 (5.53%) in RF/MW-exposed subpopulation (Table II). This gave the morbidity rates (per 100 000 subjects per year) for all age groups 97.61 for non-exposed and 178.75 for RF/MW-exposed (exposed to non-exposed ratio = 1.83) (Table II).

Table II.

Cancer morbidity in Polish career military personnel

exposed occupationally to radiofrequency and microwave radiation

- a 20 year analysis (1971 - 1990).

NUMBER OF NEOPLASMS.

Non-Exposed

Exposed

TOTAL

Size of population

120 630 ± 17900

(96.9%)

3 860 ± 770

( 3.1%)

124 500 ± 18300

(100.0%)

Number of neoplasms

(all cases)

2 355

(94.46%)

138

(5.53%)

2 493

(100.0%)

Morbidity rate

(per 100 000

per year)

97.61

178.75

100.12

Exposed/Non-exposed Ratio

1.83

Analysis of particular localizations of the diagnosed neoplasms (Table III) indicates that significantly higher morbidity rates in the RF/MW-exposed group were noted for cancers of the alimentary tract, skin tumours, including melanoma, brain neoplasms and haematologic/lymphatic malignancies. For haematologic/lymphatic malignancies the difference in morbidity between exposed and non-exposed servicemen was the largest (ratio = 5.33, Tables III and IV), although in the RF/MW-exposed group only 36 cases were diagnosed during 20 years. It should be noted that for particular types of haematologic/lymphatic malignancies (Table IV) only single cases were diagnosed in RF/MW-exposed group during the 20-year period of observation – e.g. acute lymphatic leukaemias or plasmocytomas.

Cancer morbidity in age groups (Fig.1) follows a rapidly increasing rate in subjects aged over 40 years, both in non-exposed and exposed groups. For all neoplasms the curves for cancer morbidity are parallel each other, although at the considerably higher level for the RF/MW-exposed group (Fig.1). This may suggest that in the RF/MW-exposed group the spontaneous neoplasms (or at least particular types of the neoplasms) develop faster, with shorter latency period than in non-exposed. In fact, a cumulative morbidity for a lifetime would be in this case similar, but in exposed subjects the disease occurs earlier, by 5-10 years. The only

Table III.

Cancer morbidity in Polish career military personnel

exposed occupationally to radiofrequency and microwave radiation

- a 20 year analysis (1971 - 1990).

LOCALIZATION OF NEOPLASMS.

Localization

Non-exposed

Exposed

Ex/N-Ex

ratio

No.

Rate

No.

Rate

Oral cavity

3.39

3.88

1.14

Pharynx

2.94

2.59

0.88

Esophag./Stomach

341

14.13

27.20

1.92

<0.05

Colorectal

249

10.32

18.13

1.76

< 0.05

Liver/Pancreas

3.02

3.88

1.28

Larynx/Lungs

724

30.01

34.97

1.16

Bones

2.19

2.59

1.18

Skin/Melanoma

106

4.39

9.07

2.07

< 0.05

Kidney/Prostate

146

6.05

7.77

1.28

Nervous s./Brain

3.36

9.07

2.70

< 0.01

Thyroid

2.11

2.59

1.23

Haematologic & Lymphatic

211

8.74

46.63

5.33

< 0.01

Other

167

6.92

10.36

1.49

ALL LOCATIONS

2355

97.61

138

178.75

1.83

< 0.01

exception apperas to be haematologic/lymphatic neoplasms, which show a considerable increment in number of cases in RF/MW-exposed personnel aged above 40 years (Fig.1) and for these neoplasms the depart of lines is visible at the very early age of victims.

Assessment of exposure of the investigated personnel (Table V) revealed that about 85% of the servicemen had during work shifts the average exposure levels which did not exceed 6 W/m². Correlation of cancer morbidity rates with the exposure levels (Table V), although

based on relative small number of cases (36 during 5 years) has shown a considerably higher rate in the two subgroups with high exposure levels.

A comparison of cancer morbidity rates during the first (1970 - 1979) and the second (1980 - 1989) decade and during the whole 20-year period of analysis revealed the same trends and significances for all periods, however, higher morbidity rates in RF/MW-exposed rates were noted during the decade of seventies. It is not possible to assess the exposure levels for servicemen who worked with MW fields in the fifties and early sixties, when no safety limits were yet established, however it is reasonable to assume that at that time there were numerous cases of exposure in strong MW fields and even thermal effects were noted.

To our knowledge the present data for the first time present a hint that there exist a relation between cancer risk and exposure level in RF/MW fields (Table V). There is a long way to confirm this relation, but a considerably higher morbidity rate in subjects exposed to RF/MW fields exceeding 6 W/m² clearly indicate a need to confirm this observation on a larger material.

Table IV.

Cancer morbidity in Polish career military personnel

exposed occupationally to radiofrequency and microwave radiation

- a 20 year analysis (1971 - 1990).

NEOPLASMS OF THE HAEMOPOIETIC SYSTEM

AND LYMPHATIC ORGANS.

Diagnosis

Non-exposed

Exposed

Ex/N-Ex

ratio

No.

Rate

No.

Rate

Lymphogranulo-matosis (Hodgkin)

2.15

7.77

3.61

< 0.01

Non-Hodgkin lymphoma, Lymphosarcoma

2.81

12.95

4.61

< 0.01

Chronic Lymphocytic Leukaemia (CLL)

1.49

6.48

4.35

< 0.01

Acute Lymphoblastic Leukaemia (ALL)

0.37

1.54

4.16

< 0.01

Chronic Myelocytic Leukaemia (CML)

0.95

10.36

10.90

< 0.01

Acute Myeloblastic Leukaemia (AML)

0.83

5.18

6.24

< 0.01

Plasmocytoma (PL) and other Myelomas

0.12

1.29

10.75

ALL DIAGNOSES

211

8.74

46.63

5.33

< 0.01

A comparison of own studies with other available studies on cancer morbidity in workers exposed to radiofrequency and microwave fields (for most recent review, see 3) reveals that only part of published epidemiological studies reported increased risk of certain forms of neoplasms, while in other studies no significant differences were found. However, each study, including our, suffers from certain limitations in assessment of individual past MW exposure of cancer victims and therefore, concluding about causal links is very difficult. Taking into advance the above, it is reasonable to assume at the present state of knowledge that the epidemiologic evidences still falls short of their strength and consistency required to come to a reasonable conclusion that MW can cause human cancer. Therefore, MW radiation should be still classified in group 3 (unclassifiable as to carcinogenicity in humans) of the IARC classification of human carcinogens.

Table V.

Cancer morbidity in Polish career military personnel

exposed occupationally to radiofrequency and microwave radiation

- a 5- year analysis (1985 - 1990).

EXPOSURE LEVELS AND MORBIDITY RATE

IN PROSPECTIVE STUDY (1985 - 1990).

OCCUPATIONAL EXPOSURE TO RF/MW RADIATION

Year of analysis

Percent of career personnel considered as exposed to RF/MW

AVERAGE EXPOSURE LEVELS (W/m²)

for 2 – 4 hours during working shift

(% of personnel with exposure)

1 - 2

2 - 6

6 - 10

> 10

1985

3.18%

48.2

36.6

7.9

7.3

1990

3.94%

47.3

38.1

8.3

6.3

MEAN

3.6% = 3 860 ± 770

47.8

37.3

8.0

7.1

CANCER MORBIDITY 1985 - 1990

Total number of personnel

1900

1320

350

280

Number of neoplasms (N = 36)

(38.9%)

(25.0%)

(19.4%)

(16.7%)

Morbidity rate

(per 100 000 per year)

146.9

135.8

401.4

427.0