Risk of neurological diseases among survivors of electric shocks: A nationwide cohort study, Denmark, 1968–2008

INTRODUCTION

In the beginning of the last century, when the use of electricity in both industry and the home increased rapidly, the concern for the effect of electric injuries including electric shocks on the human body increased as well. Several studies have described neurological complications of electrical injuries involving either cerebral complaints (loss of consciousness, seizures, headaches) or peripheral complaints (sensory loss, paralysis, neuropathic pain) and permanent neurological damage was the most feared complication [Chritchley, 1934; Silversides, 1964; Butler and Gant, 1977; Hunt et al., 1980; Grube et al., 1990; Arnoldo et al., 2004]. Furthermore, there have been observations of changes in the peripheral nerves such as a breaking down of the outermost layer of the nerve fibers in the peripheral nervous system as a result of electric injuries [Chritchley, 1934] and accelerated demyelination and neuronal death in brain tissue of experimental animals exposed to electric shocks [Langworthy, 1930; Morrison et al., 1930]. All these have raised the question of whether electric shocks cause neurological diseases, and studies have linked electric injuries to amyotrophic lateral sclerosis (ALS) [Haynal and Regli, 1964; Deapen and Henderson, 1986], epilepsy and Parkinson's disease [Silversides, 1964], peripheral nerve diseases [Grube et al., 1990], and vertigo [Hooshmand et al., 1989]. Simultaneously, the link between occupational exposure to electric and magnetic fields in the 50/60 Hz frequency band (extremely low frequency, ELF) and the occurrence of neurological diseases have been studied [Haynal and Regli, 1964; Deapen and Henderson, 1986; Wechsler et al., 1991; Sobel et al., 1995; Sobel and Davanipour, 1996; Feychting et al., 1998; Johansen and Olsen, 1998a; Savitz et al., 1998a, b; Johansen et al., 1999; Johansen, 2000]. However, it has been speculated that the excess risk of neurological disease found in people in occupations with above-average exposure to electric and magnetic fields may be due to repeated electric shocks associated with specific jobs, rather than to the exposure to ELF electric and magnetic fields [Johansen and Olsen, 1998a; Johansen, 2000]. (The term “electric and magnetic fields” used here is rather imprecise but the authors of the articles referred to are using the common term “ELF-EMF” without distinguishing between electric fields and magnetic fields).

To further address the hypothesis of an association between electric shocks and neurological diseases, we analyzed information for a nationwide cohort of men and women in Denmark who survived an electric accident involving an ELF electric source. By linkage and follow-up of survivors in the files of the Danish National Register of Patients (Copenhagen), the hospital discharge rates for each set of neurological diseases were calculated and compared with the respective rates for these diseases in the general Danish population, by time since date of accident.

MATERIALS AND METHODS

RESULTS

The 3,133 subjects included in the study (2,472 men and 661 women) accrued 62,982 person-years of follow-up (average, 20.1 years; range, 0–32 years; Table 1), where person-years of follow-up is the sum of the years that each subject in the cohort has been followed.

We observed significantly increased risks for peripheral nerve diseases in our electric shock cohort (SHR, 1.66; 95% CI, 1.22–2.22), migraines (SHR, 1.80; 95% CI, 1.23–2.54), vertigo (SHR, 1.60; 95% CI, 1.22–2.05), and epilepsy (SHR, 1.45; 95% CI, 1.11–1.85) when compared with the general Danish population (Table 2). The number of cases of other neurological diseases were small (all fewer than 10 observed cases), making the estimates unstable (Table 2). Most cases of neurological disease were in the group exposed to low voltages (<1,000 V), except for 12 cases of peripheral nerve diseases, migraine, vertigo, epilepsy, Alzheimer's disease, and multiple sclerosis. However, in an analysis stratified by low and high voltage, the risks for peripheral nerve diseases, migraines, vertigo, and epilepsy were still significantly increased in the group exposed to low-voltage shocks (data not shown). One person with epilepsy and one with migraines had experienced two electrical accidents before the time of diagnosis, both in the low-voltage range.

The estimated relative risks by length of follow-up showed that the risk for peripheral nerve diseases was significantly increased by almost threefold in the first 4 years of follow-up and subsequently declined to about 30% (Table 3). The risk for vertigo was significantly increased by twofold in the 5- to 9-year follow-up category, with a borderline significant increase of slightly more than 30% in the long-term follow-up. For epilepsy and migraines, the risks increased at borderline significance by 32% and 65%, respectively, in the long-term follow-up (Table 3).

DISCUSSION

We observed significantly increased risks for peripheral nerve diseases, migraines, vertigo, and epilepsy after exposure to electric shocks ranging from <50 V to more than 132 kV, with 92.3% of the shocks at voltages <1,000 V. This is the first study, to our knowledge, that set up a large-scale retrospective cohort of people who survived an accident involving exposure to electric currents and followed them for the occurrence of neurological diseases.

The peripheral nerves can be temporarily or permanently damaged by electric shocks, especially in the extremity through which the current passes [Grube et al., 1990]. Our finding of an increased risk for peripheral nerve diseases is consistent with those of earlier studies [Butler and Gant, 1977; Grube et al., 1990]. In our study, most of the peripheral nerve diseases showed up after long-term follow-up (10 years or more after the accident). Grube et al. [1990] showed that delayed neuropathy occurred in about one-fifth of the patients admitted to a burn center after injury by electric current. Examples of acute peripheral nervous system sequelae are motor loss and decreased sensation, while examples of delayed peripheral neuropathies are neuropathic pain (e.g., causalgia, severe burning pain), sensory abnormalities (e.g., paresthesia, sensation of tingling, pricking, or numbness) and motor weakness; however, sensory abnormalities can show up as both acute and delayed complications [Kobernick, 1982; Grube et al., 1990]. The highest risk in our study was seen at short-term follow-up when compared with the occurrence of peripheral nerve disease in the general population. In the study of patients admitted to the University of Washington Burn Center in Seattle, the acute peripheral neuropathy associated with high-voltage injury was likely to improve, while the symptoms of delayed peripheral neuropathy were less likely to resolve [Grube et al., 1990]. It has been reported that the median and ulnar nerves are those most often implicated in peripheral nerve injuries due to electric shocks [Butler and Gant, 1977], followed by damage to the radial nerves [Grube et al., 1990]. Peripheral nerve injuries caused by an electric shock can be due to direct thermal injury and vascular damage, which is also suggested as a mechanism for the delayed onset of neurological injury caused by an electrical injury, or a direct action on nerve function [Kobernick, 1982]. The severe burns that sometimes result from an electric shock alone can cause peripheral neurological problems. In an evaluation of 66 patients admitted for care of acute burns, 10 of which were electrical burns, 29% fulfilled the criteria for a diagnosis of peripheral neuropathy [Helm et al., 1977]. The peripheral nerves can also be stimulated even if there is no direct contact with a voltage source but only with another object within the ELF field at a different electric potential [International Commission on Non-Ionizing Radiation Protection, 1998].

Headaches have been reported as a late neurological complication of a high-voltage electrical injury [Silversides, 1964; Hunt et al., 1980] and after severe electrical injuries of the skull [Chritchley, 1934], which is consistent with our findings of an increased risk for migraines. Electrical stimulation near the dorsal raphe neurons can result in migraine-like headaches [Raskin, 1998], suggesting that an electric shock in which a current passes through the brain can cause migraines. Our observation of an increased risk for vertigo is consistent with the findings of a study in the United States of 16 patients who experienced electrical injuries [Hooshmand et al., 1989].

The physical complaints of both headache and vertigo resemble those after cerebral concussion or contusion [Silversides, 1964]. Electrical accidents may involve a fall, and injuries from a fall of considerable distance can result in severe headaches, for example. Unfortunately, our records of electrical accidents did not include information on associated falls and therefore we cannot conclude whether the observed increases in the risks for migraines and vertigo were due to the damage caused by the current passing through the body or a related fall.

In this study, in which we observed a 45% increased risk for epilepsy among people who had experienced an electric accident, we had information only on the voltage of the installation that was the cause of electrocution and could not define the actual exposure to electric and magnetic fields. We do, however, know from the records that many of the people who had an electric accident worked in jobs in which they might have had a higher exposure to electric and magnetic fields than the general population [Bowman et al., 2007; Hug et al., 2010]. Our observation is therefore consistent with the findings of a previous study of the incidence of neurological disease among employees in Danish utility companies, in which we observed a twofold excess risk for epilepsy among employees with the highest exposure to electric and magnetic fields when compared with the group with background exposure [Johansen, 2000]. In the present study, 16 of the 63 people found to have epilepsy after the electric accident worked as electricians at the time of the accident. It can be assumed that electricians, who install and repair electric circuits in factories, offices and dwellings, more frequently experience repeated electric contusions in the range of 220–380 V (230–400 V since 1994) than the general population. The same can be assumed for employees at utility companies. Therefore, the results of these two studies raise the question of whether repeated exposure to electric contusions can increase the risk of epilepsy rather than a single electric accident, although it can be assumed that much electrical installation work is carried out with the electric current turned off.

The first study linking ALS with accidental electric shocks was published in 1964, when Haynal and Regli compared the reported frequency of any previous experience of electric injuries among 73 ALS patients and 150 patients with other neurological disorders: 12 (16%) patients with ALS and 5 (3%) patients in the control group had been exposed. It was speculated that the electric trauma acted as a promoter of latent ALS. In 1986, a large case–control study (518 cases, 518 controls) by Deapen and Henderson showed a significantly increased risk for ALS, with an odds ratio of 2.8 for serious electrical shocks. However, a literature review in 2007 did not find evidence that electrical trauma is associated with an increased risk for ALS [Abhinav et al., 2007], although we found only two cases of ALS in the present study. In Denmark, the age-adjusted annual incidence of ALS is 1–2 cases per 100,000 inhabitants, with an average age at onset of 55 years [Johansen and Olsen, 1998a]. Thus, our cohort was too small to allow detection of cases of this rare neurological disease. The same applies to Alzheimer's disease, vascular dementia, Parkinson's disease, and multiple sclerosis.

Our study had several advantages, including access to detailed records of the cohort members and the general population of hospitalized cases of neurological disease identified in the National Register of Patients. Likewise, we were able to retrieve records for exposed people in the files of the Danish State Archives, the Danish Council of Electricity and the Danish Safety Technology Authority, before linking them with the National Register of Patients; therefore, bias due to selection of the study subjects is unlikely. In addition, these registries were established for administrative reasons, making information bias unlikely, and access to the unique 10-digit personal identification numbers essentially eliminates loss to follow-up or misclassification of exposure and outcomes due to erroneous linkage of register information [Olsen et al., 1995].

Our study also had limitations. Our cohort did not include all Danes who survived an accident involving an electric shock but only those cases reported to the Danish Council of Electricity or (later) the Danish Safety Technology Authority. Because all people who survived an electric accident were included in the background rates, any effect of electric shocks on neurological diseases might have been attenuated; however, our cohort probably included cases with more severe symptoms, exceeding the threshold for reporting. The small number of person-years limited our ability to investigate rare diseases of the central nervous system and precludes any conclusions with regard to ALS, for example. Another limitation is that we had no knowledge of the occurrence of diseases before January 1, 1977, but our electric shock cohort started April 1, 1968. This could attenuate the risk of neurological diseases since people in our cohort might have had a diagnosis before 1977 that we do not know of. How much this affects our results depends on the length of time between exposure and diagnosis, but there is almost no knowledge of the time span between exposure and first symptom of these neurological diseases. The identification of cases with neurological diseases from a hospital register might have introduced some degree of disease misclassification, but patients with a diagnosis of a neurological disease are normally diagnosed in neurology departments and not by general practitioners, minimizing the risk of misdiagnosed cases. We also lacked information on date of first symptoms and there might have been a delay in diagnosis when identifying cases in a hospital register. Furthermore, when patients in the early phase of the disease are not in contact with either inpatient or outpatient clinics, less severe cases of the neurological diseases studied were potentially not detected during follow-up. However, this applies to the general Danish population as well, thus leading to nondifferential disease misclassification.

Cases experiencing early symptoms of a neurological disease might have a higher risk for electric shocks because these symptoms could increase their risk for making a mistake, for example, when handling an electrical installation or machine. If this is true, the increased risks we found would be an expression of reversed causality. Nevertheless, approximately two-thirds of the cases of neurological disease were diagnosed more than 10 years after the electric accident was reported (Table 3), making it less likely that these people already had symptoms of the disease at the time of the accident.

As mentioned earlier, about 25% (16 of 63) of the people found to have epilepsy after the accident worked as electricians at the time of the accident. The same was true for 19% (6 of 32) of people with migraines, 21% (13 of 61) of those with vertigo, and 28% (13 of 46) of those with peripheral nerve diseases. Therefore, with electricians more frequently experiencing electric contusions than the general population, it is possible that frequent electric shocks might explain the increased risks for these diseases rather than a single electric accident.

In our cohort with job information available for 85% of cohort members, all kind of jobs were represented—from engineers, factory workers and fitters, to teachers, students, and housewives. Our cohort is therefore not restricted to utility workers, and the finding of a significantly increased risk for motor neuron diseases in men working in Danish utility companies compared to the general population [Johansen, 2000] does not oppose our choice of the general population as a comparison group.

In summary, in this nationwide, population-based cohort study of 3,133 survivors of an electric shock, we observed significantly increased risks for peripheral nerve diseases (66%), migraines (80%), vertigo (60%), and epilepsy (45%). Owing to small numbers, we cannot conclude whether there is an association between electric shocks and neurological diseases of the central nervous system, such as Alzheimer's or Parkinson's disease, dementia, or ALS. Our findings suggest an association between a single electric shock and an increased risk for peripheral nerve diseases, migraine, vertigo, and epilepsy; however, it might well be that repeated electric shocks were responsible for these associations. Confirmation of our results by other studies is merited.