Urbanization and traffic related exposures as risk factors for Schizophrenia

Study population

We used data from the Danish Civil Registration System [14] to obtain a large and representative set of data on Danish persons, which for all persons included current and historical information on addresses in Denmark and Greenland, and emigrations and immigrations to and from other countries together with exact dates of changes of residence. All citizens in Denmark are obliged by law to inform the authorities about any change of permanent address within 5 days. Failure to supply this information will result in inability to be on supplementary benefit (e.g., unemployment, sickness or disablement benefits, and educational aid from public funds), to be at day nursery, to be at nursery school, to attend primary and lower secondary school, to avail of free national health care, and to obtain a tax deduction card (required to have paid work), etc. Therefore, it is very unlikely that this mandatory information was not reported. This information was accessible from January 1, 1971. Our study cohort consisted of all persons with known maternal identity who were born in Denmark between January 1, 1956 and December 31, 1983, and who were alive at the 15th birthday (1.89 million persons).

Assessment of schizophrenia and mental illness in a parent or sibling

Individual information on schizophrenia in cohort members and mental illness in their relatives was obtained by linkage to the Danish Psychiatric Central Register [15]. The Danish Psychiatric Central Register has been computerized since 1969 and contains data on all admissions to Danish psychiatric inpatient facilities. From 1995 on, information on outpatient visits to psychiatric departments was included in the register. There are no private psychiatric departments in Denmark. From 1969 to 1993, the diagnostic system used was the International Classification of Diseases, 8th Revision (ICD-8) [16], and from 1994, the diagnostic system used was the International Classification of Diseases, 10th Revision (ICD-10) [17]. Cohort members were classified with schizophrenia if they had been admitted to a psychiatric hospital or had been in outpatient care with a diagnosis of the disorder (ICD-8 code 295 or ICD-10 code F20). The date of onset was defined as the first day of the first contact (in- or outpatient) with a diagnosis of schizophrenia. Parents and siblings were categorized hierarchically with a history of schizophrenia (ICD-8 code 295 or ICD-10 code F20), schizophrenia-like psychoses (ICD-8 codes 297, 298.39, 301.83 or ICD-10 codes F21-F29), or other mental disorders (Any ICD-8 or ICD-10 diagnosis), respectively, if they had been admitted to a psychiatric hospital or had been in outpatient care with one of these diagnoses.

Assessment of geographical distance to nearest major road

The Danish address and road database [13] contains information on the geographical location of all roads in Denmark combined with information on the location of house numbers on each road and with information on size of road classified in six categories: 1) motorways (110 km/h), 2) motor roads (90 km/h), 3) primary routes (80 km/h, main roads between cities), 4) secondary routes (80 km/h, main roads within cities), 5) roads 3–6 metres, 6) other roads without importance with regard to traffic structure. We used this geographical database, in SAS version 9.1 [18] and the GIS-software MapInfo Professional version 6.0 [19] to calculate 1) the geographical location of all households in Denmark, and subsequently 2) the geographical distance from each household to the nearest major road. We defined major roads by motorways, motor roads, and primary and secondary roads. The geographical distance to the nearest major road was categorized from 0 to 50 metres, from 50 to 100 metres, from 100 to 250 metres, from 250 to 500 metres, from 500 to 1000 metres, from 1000 to 2000 metres, and above 2000 metres. The completeness of information on the geographical distance to nearest major road increased from 75.5 percent in 1971 to 96.5 percent in 1978 and further to 98.5 percent in 1985.

Assessment of degree of urbanization

The 275 municipalities in Denmark were classified according to the degree of urbanization [20]: capital, capital suburb, provincial city with more than 100,000 inhabitants, provincial town with more than 10,000 inhabitants, or rural areas. Note that, the measure of the degree of urbanization used in Danish studies (E.g.[2–4, 21]) was equal for cohort members living in the same municipality, whereas the geographical distance from place of residence to nearest major road varied among cohort members living in the same municipality.

Additional File 1 shows the degree of urbanization in Denmark along with all major roads in Denmark, and Additional File 2 shows all households in Denmark coloured according to the distance to nearest major road.

Study design

Using data from the Civil Registration System, for each person in the cohort we calculated the degree of urbanization, and the geographical distance from the place of residence to nearest major road. This information was calculated for each one-year age-point from birth to the 15th birthday. Throughout this paper by place of residence we refer to place of residence at the 15th birthday, and by degree of urbanization we refer to the degree of urbanization at place of residence at the 15th birthday.

Note that, all information was independent of the disease status. A total of 1.89 million persons were followed from their 15th birthday or April 1, 1970 whichever came later, until the date of onset of schizophrenia, the date of death, the date of emigration from Denmark, or December 31, 2001, whichever came first.

Statistical analyses

The relative risk of schizophrenia was estimated by log-linear Poisson regression [22] with the GENMOD procedure in SAS version 9.1 [18]. All relative risks were adjusted for age and its interaction with gender, calendar year, history of mental illness in a parent or sibling. Age, calendar year, and history of mental illness in siblings were treated as time-dependent variables [23], whereas all other variables were treated as variables independent of time. In order to reduce the risk of residual confounding, age was categorized as, 15, 16, 17, 18, 19, 20–21, 22–23, 24–25, 26–27, 28–29, 30–34, 35–39, and > = 40 completed years. Calendar year of diagnosis was categorized in three year age bands from 1970 to 1993 and in one-year age bands thereafter. P values were based on likelihood ratio tests and 95 percent confidence limits were calculated by Wald's test [23]. The adjusted-score test [24] suggested that the regression models were not subject to over-dispersion.

Geographical distance to nearest major road at the 15th birthday and the degree of urbanization at the 15th birthday

When the effect of geographical distance to nearest major road and the degree of urbanization were adjusted mutually (Table 3, Second Adjustment), the effect of geographical distance to nearest major vanished (p = 0.77) while the effect of the degree of urbanization was not modified and remained strongly significant (p < 0.0001). This meant that the degree of urbanization explained the effect of the geographical distance to the nearest major road. Furthermore, there was no interaction between gender and geographical distance to nearest major road (results not shown).

Geographical distance to nearest major road in one-year age points and the degree of urbanization in one-year age points

Without adjusting for the degree of urbanization, the geographical distance to nearest major road from birth to the 14th birthday had an effect similar to the effect of geographical distance to nearest major road at the 15th birthday. Furthermore, when these effects were adjusted for the degree of urbanization at the same age-point, the effect of geographical distance to nearest major road at that age-point vanished while the effect of the degree of urbanization at that age-point was not modified and remained strongly significant. Therefore, the effect of the geographical distance to the nearest major road during upbringing was explained by the degree of urbanization during upbringing.

Subdividing the results above by road size (motorways, motor roads, primary routes, and secondary routes) gave similar results. Neither the effect of the degree of urbanization, nor the effect of the distance to nearest major road depended significantly on birth year.

Strengths and Limitations

Although many studies identified urban-rural differences in the occurrence of schizophrenia there is no standardized measure of urbanization; some authors defined urbanization by the density of people [25], some as the density of addresses [26–28], some by a large city vs. a small city [8, 9, 28, 29] and other definitions were also used [6, 30], and we measured urbanization by the capital area and the number of inhabitants in the largest city in the municipality [2–5, 21, 31, 32]. As long as the cause(s) or exposure(s) responsible for the urban-rural differences in schizophrenia risk are unknown we cannot give guidelines on how the degree of urbanization should be measured.

We used the distance to nearest major road as a proxy for traffic related exposures. However, we had no data documenting whether the distance to nearest major road was a reasonable proxy for traffic related exposures. Such data are currently being collected for 10,000 selected people in the capital area in Denmark. However, although the distance to nearest major road almost certainly was a crude proxy for traffic related exposures and if traffic related exposures were responsible for the urban-rural differences in schizophrenia risk, it was unlikely that the degree of urbanization would be a better proxy for traffic related exposures than the distance to nearest major road. This study suggested that traffic related exposures were not responsible for the urban-rural differences in schizophrenia risk. Previously, Mortensen [10] and Freeman [11] hypothesised that the cause(s) responsible for the urban-rural differences in schizophrenia risk might include toxic exposures, diet, infections, stress, or an artefact due to selective migration.

The strengths of this study were the use of the Danish national registers where we had information on all people living in Denmark, and that all information on exposures was measured independently of the disease.

Concluding remarks

This is the first study to investigate whether geographical distance from residence to nearest major road influences schizophrenia risk, and it is also the first study to utilise detailed geographical information on the location of each household nation-wide. Methods utilising Geographical Information Systems may prove useful in identifying the underlying unknown causes explaining the urban-rural differences in schizophrenia risk.