A study based on transformers inside buildings

What it is

Transexpo is an epidemiological study of childhood leukaemia and magnetic fields.  The central concept is that, where you have an indoors transformer in an apartment block, the apartment closest to the transformer can have a high exposure and apartments away from the transformer have low exposures.  So comparing childhood leukaemia rates between different apartments in the same building should allow you to see any effect of magnetic fields.

It is being organised as a multi-national collaboration through EPRI.

Why is this worth pursuing?

Existing epidemiological studies of childhood leukaemia have found a statistical association with magnetic fields.  But we don't know if that association reflects a true causal relationship, or whether it could be due to bias in the studies.  For example, there would be a bias in the study if the control children were systematically different from the case children in a way that affected their expoure because of the way they were selected or chose to participate.  We know that exactly this has happened in some existing studies.

Transexpo is designed to minimise this.  Firstly, a case and their control would come from the same apartment building, so there is unlikley to be a big socioeconomic difference between them.  If you start by identifying the apartment building, then work out which children live there, you help avoid selection biases as well.  And if you can deduce which apartments are exposed just from their location in the building, you won't need access or any participation by the subjects, so you avoid participation bias as well.

It's also potentially useful to look at a different source of exposure.  Existing studies have either measured fields in homes, in which case the source of the fields is usually low-voltage distribution wiring, or they have calculated the fields from high-voltage power lines.  Looking at fields from transformers, as Transexpo does, is different to either of these.  If it found the same association, that might strengthen the case for saying it's magnetic fields that matter, regardless of source; if it didn't, that might strengthen the case for it being something specific to each source.

The drawback is that in many countries, there simply aren't that many children living close to transformers in apartment buildings.  Each country would have only small numbers, and to be viable overall, the study has to be a multi-nation international collaboration.

The design of Transexpo in detail

Transexpo can have one of three different designs depending on the country-specific circumstances:

 Cohort study

  • identify from population registers all children who lived in relevant buildings
  • identify which of these lived in “exposed” apartments and which in “non-exposed”apartments
  • separately, link to cancer registry and identify which of these were cancer cases
  • compare cancer incidence rate between “exposed” and “non-exposed” children

 Nested case-control study

  • identify children from cancer registry who lived in relevant buildings
  • select control children from children who lived in relevant buildings (not necessarily the same ones)
  • identify which apartment within the building each child lived in
  • compare “exposed” to “non-exposed” ratios in cases and controls

 (This method still requires a way of identifying children who lived in relevant buildings. But it doesn’t require all children to be classified as to which apartment within the building they lived in, only the much smaller number of case and control children)

 Case-control study

  • identify children from cancer registry who lived in relevant buildings
  • for each relevant building, identify who else lived in it in the same year
  • select children from the residents of the building in that year as controls
  • identify which apartment within the building each child lived in
  • compare “exposed” to “non-exposed” ratios in cases and controls

What progress has been made?

Pilot studies in a number of countries have shown that apartments next to transformers do indeed have higher fields.

A summary of the evidence on exposures

This graph shows the evidence from six countries that suggests that the appartment immediately next to the transformer does indeed have higher exposures:

diagram showing exposures from transformers

The table below gives the numbers:

Average Exposure by Apartment Location (µT)
 NAbove Transformer / AdjacentNFirst Floor / Shares Corner or EdgeNOther Floors / Farther Away
Finland300.56 (0.17-1.55)280.21 (0.03-0.62)300.10 (0.02-0.70)
Switzerland80.59 (0.16-1.30)100.14 (0.03-0.44)30.07 (0.02-0.20)
Hungary310.98 (0.18-3.68)270.13 (0.04-0.31)300.10 (0.01-0.39)
Israel80.40 ±0.2080.080 ±0.04080.066 ±0.036
Bulgaria200.40 (0.05-2.04)170.23 (0.03-1.13)200.10 (0.01-0.62)
Netherlands110.42 (0.11-1.19)40.14 (0.07-0.28)200.07 (0.03-0.13)

Pilot studies in different countries


flag of bulgaria

Electromagn Biol Med. 2013 Jun;32(2):209-17.
Pilot study of extremely low frequency magnetic fields emitted by transformers in dwellings. Social aspects.
Zaryabova V1, Shalamanova T, Israel M. 


A large number of epidemiologic studies examining the potential effect of residential exposure to extremely-low frequency (ELF) magnetic fields and childhood leukemia have been published. Two pooled analyses [Ahlbom A, Day N, Feychting M, Roman E, Skinner J, Dockerty J, Linet M, et al. (2000). A pooled analysis of magnetic fields and childhood leukaemia. Br J Cancer. 83(5):692-698; Greenland S, Sheppard AR, Kaune WT, Poole C, Kelsh AM (2000). A pooled analysis of magnetic fields, wire codes, and childhood leukemia. Epidemiology. 11(6):624-634], which included the major epidemiologic studies on ELF magnetic fields and childhood leukemia showed twofold increase in childhood leukemia risk in association with residential ELF exposure above 0.3-0.4 μT. Based on "limited" epidemiologic evidence linking ELF exposure to childhood leukemia and "inadequate evidence" for carcinogenicity of ELF in rodent bioassays, the International Agency for Research on Cancer (IARC) classified ELF magnetic fields as a possible human carcinogen (2B classification) [International Agency for Research on Cancer (IARC) (2002). Non-ionizing radiation, Part 1: Static and extremely low-frequency (ELF) electric and magnetic fields. IARC monographs on the evaluation of carcinogenic risks to humans. Vol. 80. IARC Press: Lyon], confirmed by WHO on the basis of studies published after 2000 [World Health Organization. Extremely low frequency fields. In: 238 Environmental health criteria, Geneva: WHO; 2007]. The analysis of more recent studies of ELF magnetic fields and childhood leukemia had small findings and propose methodological improvements concerning the uncertainties in epidemiological approaches and exposure assessment, bias in selection of controls [Kheifets L, Oksuzyan S (2008). Exposure assessment and other challenges in non-ionizing radiation studies of childhood leukaemia. Radiat Prot Dosimetry. 132(2):139-147]. By the end of 2010, 37 countries had been identified for possible participation in the International study TRANSEXPO. The pilot work has been completed in five countries (Finland, Hungary, Israel, Switzerland and Bulgaria). In 2008, Bulgaria through the National Centre of Public Health Protection joined with pilot study in TRANSEXPO Project. At this first stage of the project our investigation was directed to performing measurements in dwellings with built-in transformer stations, collecting data of population and cancer registry and choosing the epidemiology design feasible for continuing the project. Taking into account the available sources of information in Bulgaria (different registers of the population) needed for epidemiological approach, it was found that the most appropriate epidemiology design would be the nested case-control study. Control group could be collected in accordance with the international requirements for such epidemiological studies. This approach could be modified in the course of the further study in order to ensure achievement of the purposes of the main international requirements of the study.



flag of finland

Bioelectromagnetics. 2008 Apr;29(3):213-8.
Indoor transformer stations as predictors of residential ELF magnetic field exposure.
Ilonen K1, Markkanen A, Mezei G, Juutilainen J.


Transformer stations in apartment buildings may offer a possibility to conduct epidemiological studies that involve high exposure to extremely low frequency magnetic fields (MF), avoid selection bias and minimize confounding factors. To validate exposure assessment based on transformer stations, measurements were performed in thirty buildings in three Finnish cities. In each building, spot measurements in all rooms and a 24-h recording in a bedroom were performed in one apartment above a transformer station (AAT), in one first floor (FF) reference apartment, and one reference apartment on upper floors (UF). The apartment mean of spot measurements was 0.62 microT in the AATs, 0.21 microT in the FF and 0.11 microT in the UF reference apartments The 24-h apartment mean (estimated from the spot measurements and the bedroom 24-h recording) was 0.2 microT or higher in 29 (97%) AATs, in 7 (25%) FF and in 3 (10 %) UF reference apartments. The corresponding numbers for the 0.4 microT cut-off point were 19 (63%), 4 (14%), and 1 (3.3%). The higher MF level in the FF reference apartments indicates that they should not be considered "unexposed" in epidemiological studies. If such apartments are excluded, a transformer station under the floor predicts 24-h apartment mean MF with a sensitivity of 0.41 (or 0.58) and a specificity of 0.997 (or 0.97), depending on the MF cut-off point (0.2 or 0.4 microT). The results indicate that apartments can be reliably classified as high and low MF field categories based on the known location of transformer stations.


J Expo Sci Environ Epidemiol. 2014 Jan-Feb;24(1):100-4.
Indoor transformer stations and ELF magnetic field exposure: use of transformer structural characteristics to improve exposure assessment.
Okokon EO1, Roivainen P1, Kheifets L2, Mezei G3, Juutilainen J1.


Previous studies have shown that populations of multiapartment buildings with indoor transformer stations may serve as a basis for improved epidemiological studies on the relationship between childhood leukaemia and extremely-low-frequency (ELF) magnetic fields (MFs). This study investigated whether classification based on structural characteristics of the transformer stations would improve ELF MF exposure assessment. The data included MF measurements in apartments directly above transformer stations ("exposed" apartments) in 30 buildings in Finland, and reference apartments in the same buildings. Transformer structural characteristics (type and location of low-voltage conductors) were used to classify exposed apartments into high-exposure (HE) and intermediate-exposure (IE) categories. An exposure gradient was observed: both the time-average MF and time above a threshold (0.4 μT) were highest in the HE apartments and lowest in the reference apartments, showing a statistically significant trend. The differences between HE and IE apartments, however, were not statistically significant. A simulation exercise showed that the three-category classification did not perform better than a two-category classification (exposed and reference apartments) in detecting the existence of an increased risk. However, data on the structural characteristics of transformers is potentially useful for evaluating exposure-response relationship



flag of hungary

Radiat Prot Dosimetry. 2008;131(4):469-73.
Exposure to 50 Hz magnetic field in apartment buildings with built-in transformer stations in Hungary.
Thuróczy G1, Jánossy G, Nagy N, Bakos J, Szabó J, Mezei G.


Exposure to 50 Hz magnetic field (MF) was evaluated in 31 multi-level apartment buildings with built-in step-down transformer stations. In each building, three apartments were selected: one apartment located immediately above the transformer room (index apartment), one located on the same floor and one on a higher floor. The mean value of measured MFs was 0.98 microT in apartments above transformers, 0.13 microT on the same floor, and 0.1 microT in on higher floors. The mean measured MF value was higher than 0.2 microT in 30 (97%) index apartments, 4 (14%) on the same floor as the index apartments and 4 (13%) on higher floors. The corresponding numbers were 25 (81%), 0 and 0, respectively, when 0.4 microT was used as cut-point. It is concluded that apartments in building with built-in transformers can be reliably classified into high and low-exposure categories based on their location in relation to transformers.


flag of israel

J Expo Sci Environ Epidemiol. 2011 Jul-Aug;21(4):365-71.
Exposure to 50 Hz magnetic fields in apartment buildings with indoor transformer stations in Israel.
Hareuveny R1, Kandel S, Yitzhak NM, Kheifets L, Mezei G.


To advance our understanding of an association between exposure to power frequency magnetic fields (MFs) and the risk of childhood leukemia, we should conduct a study that is convincingly free of selection and response bias, with highly accurate exposure assessment and a large number of highly exposed individuals. Previous measurements revealed that MF in apartments located above internal transformer stations (ITSs) are higher than in other apartments in the same building. An international epidemiologic study of childhood leukemia, TransExpo, was designed to take advantage of this scenario. This article presents the results of an exposure assessment study performed in apartment buildings with ITS in Israel. Measurements were performed in 41 apartments within 10 buildings. Average MF at the height of 0.5 m was 0.40 μT in apartments above the ITS and 0.06-0.12 μT in all other apartments. These results confirm that classification of MF exposure based on apartment location is feasible with remarkable specificity (0.98 and 0.96 for cutoff points of 0.2 and 0.4 μT, respectively) and sensitivity (1.00 for both cutoff points). Because the location of an apartment relative to the ITS can be easily determined, an exposure assessment can reliably be performed without obtaining access to residences.


Radiat Prot Dosimetry. 2012 Apr;149(2):191-5.
Time dependence of 50 Hz magnetic fields in apartment buildings with indoor transformer stations.
Yitzhak NM1, Hareuveny R, Kandel S, Ruppin R.


Twenty-four hour measurements of 50 Hz magnetic fields (MFs) in apartment buildings containing transformer stations have been performed. The apartments were classified into four types, according to their location relative to the transformer room. Temporal correlation coefficients between the MF in various apartments, as well as between MF and transformer load curves, were calculated. It was found that, in addition to their high average MF, the apartments located right above the transformer room also exhibit unique temporal correlation properties.


thumbnail flag switzerland

Sci Total Environ. 2011 Aug 15;409(18):3364-9. doi: 10.1016/j.scitotenv.2011.05.041.
Extremely low frequency magnetic field measurements in buildings with transformer stations in Switzerland.
Röösli M1, Jenni D, Kheifets L, Mezei G.


The aim of this study was to evaluate an exposure assessment method that classifies apartments in three exposure categories of extremely low frequency magnetic fields (ELF-MF) based on the location of the apartment relative to the transformer room. We completed measurements in 39 apartments in 18 buildings. In each room of the apartments ELF-MF was concurrently measured with 5 to 6 EMDEX II meters for 10 min. Measured arithmetic mean ELF-MF was 0.59 μT in 8 apartments that were fully adjacent to a transformer room, either directly above the transformer or touching the transformer room wall-to-wall. In apartments that only partly touched the transformer room at corners or edges, average ELF-MF level was 0.14 μT. Average exposure in the remaining apartments was 0.10 μT. Kappa coefficient for exposure classification was 0.64 (95%-CI: 0.45-0.82) if only fully adjacent apartments were considered as highly exposed (>0.4 μT). We found a distinct ELF-MF exposure gradient in buildings with transformer. Exposure classification based on the location of the apartment relative to the transformer room appears feasible. Such an approach considerably reduces effort for exposure assessment and may be used to eliminate selection bias in future epidemiologic studies.

The project continues to develop ways of conducting the study in the countries already involved and to explore involving other countries.

Is the UK participating?

thumbnail flag UKThe UK electricity industry investigated whether the UK could contribute as one of the nations participating in Transexpo.

To conduct a component of Transexpo in the UK, we would need:

  • a way of searching birth registers by address to generate lists of children born at particular addresses
  • a way of generating a list of all addresses either in high-rise blocks of flats (the UK equivalent of "apartment buildings"), or, better, in buildings which are served by internal substations (there are likely to be a thousand or so of these so this would ideally have to be automated)
  • a way of determining, within each building, which is the flat/apartment which is adjacent to the substation (which could probably be done by site visit as the numbers would be relatively small)

 If we could do that, we would still need to do pilot work to check that the flat next to the substation does indeed have high exposures (or, if flats next to substations in general don’t, what are the characteristics of the substations that do generate high exposures)

 At present, the two stumbling blocks appear to be:

  • there is no obvious way to assemble the lists of addresses in blocks of flats with internal substations automatically, and the numbers make it probably unrealistic to do it manually
  • if we could do the study at all, we would expect the UK study to contribute rather less than one case, (our best estimate is 0.6, on the assumption of no elevated risk).

Because the numbers appear to be too small to justify the effeort, we are not currently expecting the UK to be part of TransExpo.