Long-term exposure to electric fields. A cross-sectional epidemiologic investigation of occupationally exposed workers in high-voltage substations.

Long-term exposure to electric A cross-sectional epidemiologic investigation of occupationally exposed workers in high-voltage substations. Scand. j. work environ. & health 5 (1979) 115 125. In the present epidemiologic study, 53 workers with a long-term (more than five years) exposure to the electric field of 400 kV .substations were examined and compared with a matched reference ,group of 53 nonexposed workers from the same power companies. Matching variables included age, ,geographic location and em ployment time. The ,aim of the study was to investigate the possibility of persistent, chronic health effects in the exposed group as a consequence of exposure. The investigation included the nervous system (neurasthenic symptoms, psychological t~ts, electroencephalography), the cardiovascular system (symptoms, blood pressure, electrocardiography), and the blood (hemoglobin, red blood cells, reticulocytes, white blood cells including differential count, thrombocytes, sedimentation rate). Fertility was also assessed. The results showed no differences between the exposed and reference groups as a consequence 'of the long-term exposure to the electric fields. The groups differed, however, in that the exposed group had (a) consistently better results on the psychological performance ,tests, (b) a fewer number of children, espe cially boys, and (c) somewhat higher education. The differences in test results were due to the higher education among the exposed. The difference in number of children was also thought to be related to factors other than exposure since it was found to be present already 10-15 years before the work in 400 kV substations began.

enters an electric field may find Ithat the hair on his skin vibrates (so-called pHoerection). Spark discharges may occur between skin and clothing or between skin and grounded objects ("electric shocks"). The field intensity threshold for the emergenceof such effects is fairly well known (8). Persons wOI'king within electric fields can find the discharges disturbing.
Inadd~tion to the effects mentioned, functional disturbances have been reported in the nervous, circulatory and gastrointestinal systems of Russian workers in high-voHage substations (1,2,3,9,25,26,27,28) and in experimental studies on volunteers (17). Subjective complaints included fatigue, headache, dizziness, impaired memory, nausea, loss 'Of strength in the limbs, respiratory difficulties, increased nervousness, sleep disturbances, and reduced libido. Effects that were recorded objectively were changes in the number of blood cells, reduced systolic blood pressure, sinus arrhythmia and sinus tachycardia in electrocardiograms (ECGs), and desynchronized alpha activity and focal activations in electroencephalograms (EEGs). Impaired performance was also reported for vigilance and reaction time tests. The Russian studies covered about 500 workers at 500 'kV, 400 kV, 330 kV, and 220 kV substations. The aforementioned effects were associated with work at 400 and 500 kV substations. Work in 220 kV substations was not found to pose any heahh hazards.
These results from substation workers were 'Supported in a Spanish study on nine wOI1kers who suffered headache, dizziness, fatigue and nausea after they had begun wonk in a 400 !kV substation (10).
Results of studies on substation workers in other countries do not agree with the Russian da1a, however. Singewald et al. (30) carried out health examinations of ten linemen over a period of nine years and did not find any symptoms of nervous disorder or changes in the ECG, blood, circulatory system, lung function, kidney function, sperm pmduotion, or ,in vision or heaTing. Strumza (32) examined over 500 persons living inside and outside the immediate vicinity of 200/400 kV 1ransmission Hnes. The frequency of physician consultations, the use of medications, and the medical histories did not differ between the two groups, one living within an area of less than 25 m from the transmission lines and the other residing in an area 125 m away from the lines. In another study, no differences were found between 'a group of 84 high-voltage linemen and substation workers and a group of 94 low-voltage 'linemen with regard to a routine physical examinahon, chest radiographs, ophthalmoscopy, ECGs, and blo'Od tests (22).
The results of the health examinati'Ons, then, do not agree. The same is true with regard to experimental studies on humans (11,12,13,14,34) and animals (5, 7, 16, 116 18, 19, 20, 23, 24, 29, 31, 33, 34). As long as 25 years ago, subjectively experienced changes in taste and hearing, as well as "pains in the nerves," were reported at a field strength of 100 kV/m (14). In a laboratory experiment of more recent times, nine volunteers showed a brief, slight increase in pu'lse and blood pressure, as well as changes in GSR (galvanic skin response), upon exposure to a field strength as low as 3 kV/m (34). In contrast, no effects related to exposure to electric fields of 15 IkV/m could be noted in ECGs, EEGs, pUlse, blood ,pressure, reaction ii,me, or blood cells 'of ten test subjects (12). No effects 'on Ithe ECG, the EEG, blood pressure, or reaction time were found for six subjects exposed to 20 kV/m (11), nor did any effects show up on a psychological test or in the form of subjective symptoms when 20 subjeots were exposed to 100 kVI m (13).
It is obvious that the results available, including rtlhose from .field studies of exposed workers and those from human and animal experiments, are highly contradictory. Therefore, it is impossible to make any statement about health hazards in connection with work at substations. In fact, the effects described could not be considered to have been documented with certainty, and the individual-related exposure to electric fields was not surveyed. It was therefore deemed urgent to record individua'l exposure, wirth epidemiologic methods, Ito investigate whether long-term exposure to electric fields affects the nervous and circulatory systems or the blood and ferti'lity status of high-voltage substation wOl'kers in Sweden. Sweden began as early as 1952 to use transmission system vol,tages of 400 kV.

WORKPLACE AND WORK CONDITIONS
The examined workers were primarily employees of the Swedish State Power Board, 'but some worked for the privately owned South Swedish Power Company. The Swedish State Power Board has about 11,000 employees, about 5,000 of whom are engaged in the Operation and Administration Department.
Organizationly, the administrat10n and operation facilities are divLded geographically into five different administrative areas. The South Swedish Company has 2,000 employees, 1,600 of whom are engaged in power supply and distribution.
The personnel who work in operation and administration can be divided into two main groups, substation workers and lowvoltage distribution workers. The latter group has a comparatively insignificant exposure to electric fields.

High-voltage substation workers
Substation workers in 400 kV plants can be div'ided into two subgroups, namely, station personnel and circuit breaker personnel.
Station personnel are placed permanently at one or sevel'al neighboring sUbstations, where they carry out the everyday tasks of maintenance and operation. The work is organized differently within the different administrative units. At some of the substations, a work ,rotation system is used, whlle 'at others the personnel are divided into a maintenance group and an operation group. Operation is monitored by the person who has the so-caned "check-up" week. A "check-up" week, which occurs about once a month per worker, calls for making readings from the instruments, writing reports, recording statistics and ma:king inspection rounds throughout the substation. If the substation is unmanned at night, the person is on duty at home where the news of an alarm may reach him via telephone. Making the rounds entails a visual inspection of the equipment throughout the entire plant at the same time that certain readings are being taken from the instruments. The munds last between 1.5 to 2.0 h and take place two or three il::imes per week. Other common tasks in the substation are switching, correcting defects, cleaning the brea:kers, painting, changing light bulhs, mowing the lawn, shoveling snow, etc.
The circuit breaker personnel are not stationed at any special substation. Instead, they work wilthin an entire administrative area. They often work away from home since the administrative areas are large. Their main tasks are the inspection and maintenance of breakers and disconnecting switahes. To a ·smaller degree, they wOl'k with tap changers on transformers. Maintenance measures are carried out on breakers and disconnecting switches according to a certain routine. The wor,k time for, e.g., making revisions in a breaker varies between 2 and 7 d, depending upon the type of breaker involved.

Low-voltage distribution workers
The reference group was selected from this personnel category. The low-voltage distribution personnel work chiefly with a voltage of 220/380 V. Occasionally tasks occur involving exposure to system voltage of up to 20 kV. The work, which can generally be done without travel away from home, includes, e.g., the changing of insulator chains and conductors, the digging of cable trenches, inspection and maintenance of cable terminal cabinets, construation of new transmission lines and stations, and meter-reading at the consumers.

Exposure to electric fields
Exposures were determined in 20 substations in different parts of the country. The subsequent evaluation was made wirth respect to {a) field strengths at a height of 1.8 m in the substations and (b) exposure during different parts of the work routines.

Field strengths
Field intensirties at a height of 1.8 m in the substations were measured with a field intensity meter, which measures the undisturbed field (19). As an example, it may be mentioned that so-called highbuilt 'Substations m'ay have large areas with field intensities up to 10-15 kV/m, while so-called low-built substations may have substantiall areas with field intensities of about 20 kV/m. ladders or scaffoldings. These put the worker on the same level as the breaker, i.e., at ,a height 'Of 6-8 m. The field intensity art these places is considerably different from the intensities measured at ground -level because the grounded breakers provide partial shielding there. The degree of exposure in breaker work depends, then, a grea,t deal upon how often the worker is on an elevated level beside the breakers and how close he is to the live pallts. Table 2 shows the average exposure levels for two different breaker operations in a 400 kV substation, namely, revision and testing. The table shows that the exposure is considerably higher in the former than in the latter. This difference stems from the fact that revisions are made to a substantial degree while the worker is in ,an elevated position, whereas testing is carried out almost entirely on the ground.  Exposure during different parts of the work routine

EXAMINED GROUPS
The exposure was recorded as the actual work routines were being carried out. The person being ,tested had a portable dosimeter which registered the number of seconds he spent.in field strengths above three given levels.
Station personnel. Table 1 shows the average -exposure levels for the person maJking the rounds (visual inspection and instrument readings) and for the person doing general wonk involving large areas of the 'Substa,tion, e.g., mowing the lawn. The table show'S that the exposure for carrying out Ithe general tasks at the 'Substation is somewhat higher than that for making the rounds. This difference stems primarily from the fact that the person making the rounds caHies out the instrument readings and inspection routines as he st'ands beside grounded cable racks which shield him from the electrk field.
Circuit breaker personnel. A substantial part of the w<Jl'k involved in the repair or revision of brea-kers is performed from

Selection
During the spring -of 1974, the various regiona'l administrations were informed about the planned study, and personal information was collected on potential subjects. Information was obtained on 178 high-valtage 'Substation workers for the exposed group all!d -on 350 low-voltage distl.'libution wOl.'lkers for the reference group. For each substation worker, a referent was -chosen who was as similar as possible to the exposed person with respect to geographic location, age, duration -of employment at f1;he Swedish State Power Board/South Swedish Power Company, and also, whenever possible, with respect to duration of employment in electric production in general. A special questionna,ire coneerning the period of empl}oy1ffient in 400 kV plants was sent to a,ll of the 178 exposed persons. The exposed group was chosen so as :to inclU!de substation workers (station personnel and circuit breaker personnel) who had . _ ---_..._ --------    Table 5 shows that, on the average, the exposed group had a higher education than the reference group. By a basic education beyond the compulsory education in Sweden (previouslyelementary school up to and including the siJQth grade; now elementary school plus junior high school up to and including the ninth grade) is meant advanced studies at a vocational school, ,an apprenticeship school, or the li:ke. Noone in either of the groups had a high school (college preparatory) Exposed group -reference group worked more than five years in 400 kV plants as of the beginning of 1975, a total of 68 persons. During the fall of 1975, a questionnaire concerning education, marital status, number of children, housing conditions, dietary habi!ts, use of tobacco and alcoholic beverages, medical history, history of accidents, and use of medications was distributed. The questionnaire was filled out by all hut two of Ithe exposed and two of the referents.
The medical ·exa'minations were carried out in different parrts of the country during the period between March and October 1976. A total of four of the exposed and eight of the referents did not undergo the examination. The reasons for their dropping out are shown in table 3.
Furthermore, it turned 'out, during subsequent investigations, Ithat three of the referents had been exposed, and they were therefore also excluded. Every time an exposed or a referent dropped out, his respective referent or exposed match was also excluded. The final material thus came to consist of 53 matched pairs.
The country was divided geographically into three 'Parts: northern, central and southern Sweden. The test subjects and referents were distributed in the following manner: 13 pairs from northern Sweden (Norrbottenand Central Norrland regional administrations), 27 pairs from central Sweden (.A:lvkarleby and Motala regional administrations), and 13 pairs from southern Sweden (Trollhatten regional administration and South Swedish Power Company).
Age. At the beginning of 1975, the average age was 46.0 years for ,the exposed group and 45.8 years for the reference group. The ,age distribution is given in table 4. -------------Length of employment. At the beginning of 1975, the average ,length of employment was 21.0 years in the exposed group and 15.6 years in the reference group. It should be noted that a number of the persons in the reference group had previously been employed by smaller companies, which had later been taken over by the Swedish State Power Board. used by us in other epidemiologic investigations (15). education. With regard to employer-sponsored courses, only those exceeding one month were included. Table 6 shows that the exposed group, on the average, was away from home more often during the week than the reference group was.

Special living conditions.
Use of alcohol and tobacco. No differences in alcohol or tobacco consumption were found between the groups. The average alcohol consumption was 164 gl month in rthe exposed group and 161 gl month in the reference 'group. {In the calculations, the content of akohol was assumed to be 32 .gIl for beer, 120 gil for wine, and 334 gil for hard liquor.) The average tobacco consumption was 55 gl week in the exposed group 'and 53 glweek in the reference group.
Live at home 37 Live away from home 11 Sometimes at home and sometimes away 5 Electroencephalography. The EEGs were recorded by means of an 8-channel Siemens-Elema EEG machine. The subjects were awake in a semireclined position with their eyes closed during the 'recording. The electrodes were placed according to the 10/20 system. The paper recordings were eva'luated in the following way: (a) They were visually examined according to a rourtine clinical procedure. (b) From each EEG a typical 10-s secHon (8 channels) was selected. These sections from each and everyone of the persons examined (2 X 53) were ranked without any knowledge of which subjects they belonged to. This ranking procedure was carried out so that the EEGs showing a distinct, staMe alpha aclivi,ty were .given low ranks, while higher ranks were assigned to EEGs with ,a decreasing amount of rhythmic activity. This procedure was recenHy described in an epidemiologic study on jet fuel-exposed industrial workers (15).
Psychological tests. Eight performance tests (6,15) were carried out individuaHy by each subject. The subjects were studied in such an order as 11:0 obtain comparability between the exposed and reference groups with regard to time of day and d'ay of week for the examination. The tests were perfol'med by each subject in the following Living conditions

Nervous system
Circulatory system Standardized interviews. Data on the prevalence of neurasthenic symptoms, anxiety and depression, frequency of physician consultations due to such symptoms, and data on confounding and effect-modifying factors were collected with the use of a modified version of a standardized interview worked out ,at .'.1 Scandinavian meetingon health hazards connected with the use of solvents (4). The interview has been The examination included standardized interviews on symptoms and signs, ECG (standard 'leads from extremities and six chest leads), standardized blood pressure measurements (on the left arm after 15 min in a supine position). Codlings and inter.pretations of the results were made by a physician without knowledge of whether the subject belonged rto I1:he exposed or nonexposed group. The ECGs were classified according to the Minnesota Code.

Blood
The blood tests included analyses of hemoglobin, hematocrit, number of red blood cens, reticulocytes, white blood cells (with differential counts), thrombocytes and sedimentation rate.

Fertility
Fertility was assessed from the data collected on the number of children born, their sexes, and the dates of birth in relation to exposure, Le., if they occurred before or after the dates on which the fathers began to wOIik at high-voltage substations.

Nervous system
Standardized interviews. No differences were found -between the exposed and reference groups with regard to responses to quesbions on neurasthenia or confounding and effeot modifying famors. For instance, 16 persons in the exposed group and 17 of the refenmts responded "yes, often" to one or several of the neurasthenia questions of the interview.
Psychological tests. As can be seen in table 7, the average performance of the exposed group was, except for the Simple RT rtest, consisten1Jly better than -that of the reference group. In five out of ten comparisons the differences between the groups were statisticalily significant (p < 0.05).
When the results were subjected to further analysis, it became apparent that the differences in performance between the two groups could be ascribed almost entirely to the differences in educational background. On one hand, it was found that the differences in performance on the various tes<ts were very small when exposed wOI1kers were compared with nonexposed workers of the same educational baokground, and, on the other, the differences were greatest when workers with different educational backgrounds were compared irrespective of whether they be- Table 7. Mean values and standard deviations of the performance on psychological tests for the exposed and reference groups and p-values from the two-sided signifance testing of the differences between the mean values. As mentioned earlier, the history of signs and symptoms from the heart was recorded according to a standardized interview/ questionnaire, and the ECG results were classified according to the Minnesota Code. In the exposed group, 36 persons were therefore classified as normal versus 30 in the reference group. No statistically significant differences were found between the exposed and referents with respect to the number of persons with various heart symptoms.
The number of persons with hypertension totaled 8 in the exposed group and 11 in Ithe reference group. The systolic and diastolic blood pressure was, on the average, of the same order of magnitude in the exposed group (142/92) and in the control group (145/94). normal pattern, mainly focal abnormalities. There was no difference between the groups when the EEGs were ranked with regard to rhythmic alpha activity.

Blood
The results of ,the blood analyses did not reveal a systematic difference between the exposed and reference groups. Fertility Fig, 1 A-C. Cumulative number of children (A), boys (E) and girls (C) of the exposed (solid line) and reference groups (dotted line) over time, "0" on the time axis denotes real and hypothetical "beginning of exposure" of the exposed and reference groups, respectively. Insets: statistical fourfold tables, longed to the exposed or the nonexposed group, Electroencephalography. A visual inspection of the EEGs showed 11 of the exposed and 14 of the referents Ito have slight or moderate deviations from the The feDtility in the exposed and reference groups was 'assessed from data collected on the number and birth dates of children (boys/gil'ls) of the examined wOl'kers. In the exposed group children were born both before and after their fathers began to work in high-voltage substations, Le., before and after the start 'of exposure to electric fields. In order to make comparisons with the reference group possible, we assigned each .af the matched referents a hypothetical "start year for exposure" which agreed with that .af the relevallit WOI1ker in the exposed group. Fig. 1 A shows the cumulative number of childbirths per year up to the beginning of 1975. The year thai exposure began is indicated with "0" on the time axis. The diagram and the fourfold table within the diagram show that a total of 79 children were born in the exposed group versus 116 in the referent group. This difference is statistically significant (;(2 = 7.02; P = 0.008).
It appears from the diagram that the difference between the exposed and reference groups (79 and 116 children, respectively) arose successively, and it started long before the start of exposure. There is thus no statistically significant relationship between exposure and the number of children born in the two groups (;(20 .48; P = 0.488). Fig. 1 Band C show the cumulative number of boys and girls, respectively. The diagrams show that the difference between the two groups as to number of children has mainly to do with male births (38 and 65 boys, respectiv€ly, ;(2 = 7.08; P = 0.008). The difference with regard to female births is less pronounced and not statistically significant (41 and 51 girls, respectively, X 2 = 1.09; P~0.296). No relationship between exposure and number of childbirths could be found when the materia'! was divided according to sex (boys/girls) (;(2 =0 1.55; P = 0.213 and X 2 .~0.02; p~0.888, respectively).

DISCUSSION
In the present comparative investigation the exposed workers, on the average, were found to have (a) fewer children, espe6aHy boys, (b) better performance results on tests of psychomotor and cognitive functions, and (c) a higher education than the referents.
The exposed and referent groups were not equivalent, thus, as regards education, which is probably the explanation for the differences in the psychological test results between th€ groups. The group differ€nces in education and ,test results 'also indicate that the groups were not fully comparable as to socioeconomic and personality variables. Against such a background the ob-served difference in number of childbirths cannot be regarded as surprising -all the more as this difference developed and existed long before the beginning of exposure ·and was relatively constant during the entire length of the observation time ( fig. 1). It should be pointed out that the differences between the groups with respect to the ratio boys/girls cannot be attributed to differences in back,ground faotol's. The possibility cannot be excluded that work in electric fields played some role in this result. However, we consider it unlikely in light of the overall assessment of the results of the study. It should ·also be pointed out that, if work in substations, contrary to what seems to be the 'case in~his study, had affected the number of childbirths and/or the ratio boys/girls, the effect could be ascribed, hypothetically, just as well to other important aspects of the work, such as stress, awareness of accident risks, discomfort caused by fear of discharges, etc.
The resul1ts of the present study show no eV'idence for the development of persistent, chronic health effects in high-voltage substation workers as a consequence of exposure to electric fields. Thus ,the Russian and Spanish findings (see Introduction) were not confirmed by the present investigation. There are many possible reasons for Ithe discrepancy in results, but it should be recognized that in most of the earlier investigations on substation workers there were severe limitations, e.g., the use of too few subjects and/or the lack of a reference group.