Burden of disease due to transportation noise in the Nordic countries Gunn Marit Aasvang 1 Norwegian Institute of Public Health Department of Air Quality and Noise, Lovisenberggata 8, 0456 Oslo, Norway Centre for Disease Burden, Zander Kaaes gate 7, 5015 Bergen, Norway Leo Stockfelt 2 University of Gothenburg Occupational and Environmental Medicine, School of Public Health and Community Medicine, Institute of Medicine, Sahlgrenska Academy, Medicinaregatan 16A, 405 30, Gothenburg, Sweden

Sahlgrenska University Hospital Department of Occupational and Environmental Medicine, Medicinaregatan 16A, 405 30, Gothenburg, Swe- den. Mette Sørensen 3 Danish Cancer Society Research Center Diet, Genes and Environment, Strandboulevarden 49, 2100 Copenhagen, Denmark Anu Turunen 4 Finnish Institute for Health and Welfare P.O. Box 95, FI-70701 Kuopio, Finland Nina Roswall 5 Danish Cancer Society Research Center Diet, Genes and Environment, Strandboulevarden 49, 2100 Copenhagen, Denmark Tarja Yli-Tuomi 6 Finnish Institute for Health and Welfare P.O. Box 95, FI-70701 Kuopio, Finland Mikael Ögren 7 University of Gothenburg Occupational and Environmental Medicine, School of Public Health and Community Medicine, Institute of Medicine, Sahlgrenska Academy, Medicinaregatan 16A, 405 30, Gothenburg, Sweden

Sahlgrenska University Hospital Department of Occupational and Environmental Medicine, Medicinaregatan 16A, 405 30, Gothenburg, Swe- den.

1 GunnMarit.Aasvang@fhi.no 2 leo.stockfelt@amm.gu.se 3 mettes@CANCER.DK 4 anu.turunen@thl.fi 5 roswall@cancer.dk 6 tarja.yli-tuomi@thl.fi 7 mikael.ogren@amm.gu.se

worm 2022

Virpi Kollanus 8 Finnish Institute for Health and Welfare, P.O. Box 95, FI-70701 Kuopio, Finland Timo Lanki 9 Finnish Institute for Health and Welfare, P.O. Box 95, FI-70701 Kuopio, Finland

University of Eastern Finland Yliopistonranta 1, 70210 Kuopio, Finland . Jenny Selander 10 Karolinska Institutet Institute of Environmental Medicine, Nobels väg 13, 171 77 Stockholm, Sweden Natalia Vincens 11 University of Gothenburg Occupational and Environmental Medicine, School of Public Health and Community Medicine, Institute of Medicine, Sahlgrenska Academy, Medicinaregatan 18A, 41390 Gothenburg, Sweden Andrei Pyko 12 Institute of Environmental Medicine, Karolinska Institutet Nobels väg 13, Karolinska Institutet, Solna, 17177 Stockholm, Sweden

Center for Occupational and Environmental Medicine, Region Stockholm Solnavägen 4, plan 10, 113 65 Stockholm, Sweden Göran Pershagen 13 Institute of Environmental Medicine, Karolinska Institutet Nobels väg 13, Karolinska Institutet, Solna, 17177 Stockholm

Center for Occupational and Environmental Medicine, Region Stockholm, Solnavägen 4, plan 10, 113 65 Stockholm, Sweden Gerhard Sulo 14 Norwegian Institute of Public Health Centre for Disease Burden, Zander Kaaes gate 7, 5015 Bergen, Norway Anette Kocbach Bølling 15

Norwegian Institute of Public Health Department of Air Quality and Noise, Lovisenberggata 8, 0456 Oslo, Norway Centre for Disease Burden, Zander Kaaes gate 7, 5015 Bergen, Norway

8 virpi.kollanus@thl.fi 9 timo.lanki@thl.fi 10 Jenny.Selander@ki.se 11 natalia.vincens@amm.gu.se 12 andrei.pyko@ki.se 13 Goran.Pershagen@ki.se 14 Gerhard.Sulo@fhi.no 15 Anette.Kocbach@fhi.no

worm 2022

ABSTRACT Environmental noise is of considerable public health concern, and quantification of the health im- pacts is important for preventive strategies. We estimated the burden of disease (BoD) due to road traffic and railway noise in four Nordic countries and their capital cities in terms of DALYs (Disa- bility-Adjusted Life Years). Available data on noise exposure were used, including data from strategic noise mapping according to the Environmental Noise Directive, END (Directive 2002/49/EC). High degree of noise annoyance (HA), high degree of sleep disturbance (HSD) and ischemic heart disease (IHD) were included in the main analyses based on exposure-response associations recommended by WHO. Country-specific estimates of IHD from the Global Burden of Disease (GBD) study were used. Since several aspects of the noise exposure modelling vary considerably between the Nordic countries, no comparable estimates could be made for the entire countries. For the capital cities comparable estimates ranged from 330 to 485 DALYs/100 000 for road traffic noise, and from 40 to 140 DALYs/100 000 for railway noise. High annoyance and high degree of sleep disturbance ac- counted for the largest part of DALYs for road traffic and railway noise, respectively. Further har- monization of noise exposure modelling is important for comparative disease burden assessment. 1. INTRODUCTION

Environmental noise is a public health problem as it leads to annoyance, sleep disturbance and in- creased risk of cardiovascular disease [1, 2]. In the last decade, there has been a continued scientific effort to study the long-term effects of environmental noise, revealing an increasing body of evidence for associations also with diabetes and stroke, especially for road traffic noise [3]. Environmental noise is estimated to be the second most important environmental risk factor, after air pollution, in terms of disease burden in the EU [4]. Thus, it is important to further develop harmonized methods to reliably quantify the burden of disease (BoD) of environmental noise, at the regional, national, and global level to assist policy makers engaged in developing prevention strategies. The objective of this paper is to estimate BoD due to road traffic and railway noise in the Nordic countries using a harmo- nized approach. We aimed to employ data from the noise mapping according to END as the primary source of exposure data. Finally, we compared BoD estimates based on noise exposure according to END versus nationwide noise models. 2. METHODS

Burden of disease (BoD) due to road traffic and railway noise was assessed for each of the Nordic countries: Norway, Finland, Denmark and Sweden and for their respective capital cities: Oslo, Hel- sinki, Copenhagen, and Stockholm. BoD was assessed in terms of DALY (Disability-Adjusted Life Years), which is the sum of the number of years lost due to ill-health/disability (YLD) and years of life lost (YLL) due to premature death.

We reviewed all available national and sub-national data on noise exposure assessments for road traffic and railway, focusing on data from strategic noise mappings in accordance with the Environ- mental Noise Directive, END (Directive 2002/49/EC) for the year 2016, both those that were reported to the European Commission (EC) and those that were available in other reports. As strategic noise mapping according to END for L den ≥ 55 dB and L night ≥ 50 dB only includes agglomerations with more than 100 000 inhabitants and larger roads and railways, we also included nationwide noise map- pings where available, i.e. for Denmark and Norway. These nationwide models cover all residential buildings in the country. We used data for traffic noise from these nationwide models for the year closest in time to the strategic noise mappings according to END from 2016.

worm 2022

High degree of annoyance (HA), high degree of sleep disturbance (HSD) and ischaemic heart disease (IHD) were included as the health outcomes in the main BoD analyses based on exposure-response functions from the WHO systematic reviews [1-3]. Estimation of DALYs due to stroke and diabetes attributable to road traffic noise was included in additional analyses, using risk estimates from a recent large Nordic study [5] and a project specific meta-analysis (not reported here); these were 1.06 (95 % CI: 1.03 -1.08) and 1.11 (95 % CI: 1.04 -1.18), respectively. Country-specific YLD and YLL estimates from the Global Burden of Disease (GBD) study were used as background health data for IHD, stroke and diabetes. For HA and HSD we used disability weights (DW) as recommended by WHO [6].

In addition to assessing estimates for each health outcome separately, the estimates were summa- rized to arrive at total DALY estimates. When summarizing the YLD/YLL estimates from the various health outcomes associated with exposure to noise, any overlap between these outcomes will result in double counting. The highest dependency was considered to be between HA and HSD and based on an evaluation of overlap between responses on standard noise annoyance and sleep disturbance questions from previous Norwegian and Finnish surveys, we quantified source specific correction factors. Based on the magnitude of the overlap, we employed a correction factor of 0.8 for HA due to road traffic noise and 0.75 for HA due to railway noise, before summarizing with HSD and the disease specific BoD estimates.

The DALY estimates are reported as rates (DALYs/100 000 inhabitants) since the population sizes differ between the Nordic countries and capital cities. The estimation of 95% confidence interval (CI) was based on the 95% CI for the risk estimates, except for HA for which we calculated an uncertainty interval based on the 95% CI for HSD. 3. RESULTS

3.1. DALYs attributable to road traffic and railway noise for the Nordic capital cities The noise exposure data reported to European Commission (EC) according to END for the Nordic countries differed with respect to the applied geographical definitions for agglomerations, for instance in terms of inclusion of adjacent agglomerations in the capitals. Moreover, the noise exposure as- signment method differed, as some countries used the noise level assigned to the most exposed façade while others used the equal distribution principle. For the Nordic countries, comparable noise expo- sure data were only identified for the capital municipalities and for the most exposed façade assign- ment method. The resulting DALY rates (DALY/100 000) for road traffic noise summarized across the three health outcomes varied from 329 (95% CI: 147- 498) in Stockholm to 485 (95% CI: 232- 719) in Oslo (Table 1). The corresponding numbers for railway noise were considerably lower, rang- ing from 44 (95% CI: 24-64) in Stockholm to 146 (95% CI: 82-213) in Oslo (Table 1), and also showed more variability between the Nordic capital cities than for road traffic noise.

HA and HSD contributed the major part of the estimated DALYs attributable to road traffic noise for the capital municipalities, in total between about 80 and 90% (data not shown). Inclusion of the additional outcomes stroke and diabetes increased the total DALY estimates by 11-16% (Table 1). For railway noise, high sleep disturbance (HSD) contributed with the highest estimate, with approx- imately between 64 and 70 percent (data not shown).

worm 2022

Table 1. Estimated DALYs/100 000 inhabitants due to road traffic and railway noise in the Nordic capital municipalities. Estimates are based on the main analyses, that included HA, HSD and IHD (road traffic only), and for the analyses including the additional health outcomes stroke and diabetes (road traffic only). The esti- mates are based on the most comparable data for the capital municipalities.

Population DALY/100 000 1 DALY/100 000 1

Capital municipalities

Main outcomes (HA, HSD, IHD)

All outcomes (Main outcomes + stroke and diabetes)

Road traffic noise DALY (95% CI) DALY (95% CI) Oslo 658 390 485 (232 - 719) 543 (258 - 804) Helsinki 620 715 405 (173 - 619) 472 (203 - 719) Copenhagen 591 481 447 (217 - 657) 497 (240 - 731) Stockholm 911 989 329 (147 - 498) 379 (169 - 572)

Main outcomes (HA, HSD)

Railway noise

DALY (95% CI)

Oslo 658 390 146 (82 - 213) Helsinki 620 715 118 (65 - 173) Copenhagen 2 539 018 58 (32 - 85) Stockholm 911 989 44 (24 - 64)

1 Corrected for overlap between HA and HSD. 2 DALYs are based on the Danish nationwide noise model, 2015 as no railway noise data based on END is available for Copenhagen, 2016.

3.2. DALYs based on noise mapping according to END versus nationwide assessments To explore to which extent the geographical coverage affects the national DALY estimates, results based on END and the two nationwide models were compared, for Norway and Denmark. For road traffic noise, the DALY/100 000 for Norway was 51% higher when based on data from the nation- wide model compared to END. For Denmark, the DALY rates were more than doubled (+133%) for the nationwide model compared to END. 4. DISCUSSION

In the present BoD analyses we aimed at employing similar input data and a harmonized approach to obtain comparable DALY estimates for the Nordic countries and capital cities. This was however hampered by differences in noise exposure assessment between the Nordic countries, primarily in terms of differences in definition of geographical areas and the noise exposure assignment method. In line with this, a recent health impact assessment for European cities also revealed several differ- ences in the strategic noise mapping methods conducted by the European countries according to END and emphasized the need for more harmonized and comparable noise exposure data [7].

The capital municipalities were found to be most comparable in terms of noise exposure data, and the resulting BoD estimates showed the highest burden attributable to traffic noise exposure in Oslo and the lowest in Stockholm. The estimated burden was about 50% higher in Oslo than in Stockholm, although the uncertainty intervals were overlapping. Whether these differences in DALY estimates reflect additional differences in noise assignment methods or actual differences in disease burden due to noise in the Nordic capital municipalities, was not possible to determine within the scope of the

worm 2022

current project. A higher attributable burden due to road traffic in Oslo compared to the other Nordic capital cities was also reported in the recent study by Khomenko et al. [7] using modelled exposure data partly based on the strategic noise mapping according to END.

A direct comparison with previously reported BoD estimates is not feasible due to differences in methodological aspects, e.g., choice of included health outcomes, exposure-response functions and background health data. Moreover, except from the recent study by Khomenko et al. [7], there is a lack of studies including several Nordic capital cities. However, our finding that the main disease burden due to transportation noise is from annoyance and sleep disturbance across the Nordic capitals, is in line with previous findings [6, 8].

We found considerable differences between the estimated DALYs based on the nationwide com- pared to the END based data. While the nationwide noise exposure models cover all residential build- ings in the country, END only includes inhabitants in agglomerations with more than 100 000 inhab- itants and along major roads. We acknowledge that there are other differences between the nationwide and END models than national coverage that will be further explored in follow up analyses. The presented DALY estimates should be interpreted with caution, as there will always be uncertainties in quantitative assessments of health impact. These uncertainties relate to the exposure data, the ex- posure-response functions, the disability weights as well as the baseline health and population data. Although harmonization of input data has several strengths e.g., the use of country specific YLD and YLL estimates from the GBD study, the use of common exposure-response functions may underes- timate or overestimate the burden depending on the representativeness of these associations for the different areas and populations. Although a common Nordic noise prediction model is developed for the Nordic countries, different assumptions and choices are made by the individual countries, e.g., for the façade assignment method and when linking noise exposure to population data.

Furthermore, we introduced a correction factor for high annoyance assuming a certain overlap between reported annoyance and sleep disturbance due to noise. This is however, based on limited data, and further studies are needed to be able to summarize the outcome specific YLDs to total DALYs to avoid possible double counting.

Finally, the burden of disease due to transportation noise is estimated for noise levels of L den ≥55 and L night ≥ 50 dB. Since a certain proportion of the population will experience high noise annoyance, high degree of sleep disturbance, as well as other adverse health outcomes, also below these noise levels, these cut-offs probably contribute to an underestimation of the attributable DALYs. 5. CONCLUSIONS

This is the first study on BoD due to transportation noise in the Nordic countries and capital cities aiming at employing a harmonised approach using comparable input data. Comparison of the availa- ble exposure data across the Nordic countries revealed that the noise exposure assessment performed according to the Environmental Noise Directive, END (Directive 2002/49/EC) was not harmonized. Comparable noise exposure data were only identified for the Nordic capital municipalities. Thus, more work is required to harmonise the noise exposure data between countries, which is a prerequisite for comparative disease burden assessment.

The estimated DALYs based on available nationwide noise exposure assessments for Denmark and Norway strongly suggest that application of noise exposure data according to END leads to an underestimation of the burden due to transportation noise at the national level.

5. ACKNOWLEDGEMENTS

We gratefully acknowledge NordForsk for funding the NordSOUND study (grant number 83597).

worm 2022

6. REFERENCES

1. Guski, R., D. Schreckenberg, and R. Schuemer, WHO environmental noise guidelines for the European region: A systematic review on environmental noise and annoyance. International Journal of Environmental Research and Public Health, 14 (12), 2017. 2. Basner, M. and S. McGuire, WHO environmental noise guidelines for the european region: A systematic review on environmental noise and effects on sleep. International Journal of Environmental Research and Public Health, 15 (3), 2018. 3. van Kempen, E., et al., WHO environmental noise guidelines for the European region: A systematic review on environmental noise and cardiovascular and metabolic effects: A summary. International Journal of Environmental Research and Public Health , 15 (2), 2018.. 4. European Environment Agency, Healthy environment, healthy lives: how the environment influences health and well-being in Europe. EEA Report NO 21/2019. Luxembourg: Publications Office of the European Union. 2020. 5. Roswall, N., et al., Long-Term Exposure to Transportation Noise and Risk of Incident Stroke: A Pooled Study of Nine Scandinavian Cohorts . Environ Health Perspec t, 129 (10): p. 107002, 2021.. 6. WHO, Burden of disease from environmental noise. Quantification of healthy life years lost in Europe . WHO Regional office for Europe: Copenhagen, Denmark, 2011. 7. Khomenko, S., et al., Impact of road traffic noise on annoyance and preventable mortality in European cities: A health impact assessment. Environ Int, 162 : p. 107160, 2022. 8. Eriksson, C., T. Bodin, and J. Selander, Burden of disease from road traffic and railway noise- a quantification of healthy life years lost in Sweden. Scandinavian Journal of Work, Environment and Health, 43 (6): p. 519-525, 2017.

worm 2022