Years of Life Lost to Death
A comprehensive analysis of mortality in Germany conducted as part of the BURDEN 2020 project
; ; ; ; ; ;
Background: Knowing which diseases and causes of death account for most of the years of life lost (YLL) can help to better target appropriate prevention and intervention measures. The YLL in Germany for specific causes of death were estimated as part of the BURDEN 2020 project at the Robert Koch Institute.
Methods: Data from cause-of-death statistics were used for the analysis. ICD codes were grouped into causes of death categories at different levels of disaggregation. The YLL were estimated by combining each cause of death with the remaining life expectancy at the age of death. Deaths and YLL were compared by sex, age category, and regional distribution.
Results: Approximately 11.6 million years were estimated to be lost in Germany in 2017, of which 42.8% were lost by women and 57.2% by men. The largest number of YLL were due to (malignant) neoplasms (35.2%), followed by cardiovascular diseases (27.6%), gastrointestinal diseases (5.8%), and neurological diseases (5.7%). Deaths at younger ages had a greater impact on population health if expressed in YLL: the death share of persons under age 65 was 14.7%, but the years of life lost in this age group amounted to 38.3% of all YLL. The most common causes of death in this group include accidents, self-injury and violence, malignant neoplasms, and alcohol-related diseases.
Conclusion: A large proportion of YLL is borne by young and middle-aged persons. These findings emphasize the need to introduce preventive strategies early in life to reduce the YLL at younger ages, as well as to prevent risk factors for diseases in older ages.
Burden of disease analyses are being conducted across the world in order to map the health status of populations comprehensively and comparably according to a standardized concept (1, 2). One important element of such analyses is disability-adjusted life years, a population-based composite measure that combines mortality and morbidity to create an indicator for the health of the population concerned (3). The focus is not on the number of cases of disease and death; rather, the actual impact on health is described in a combined unit of measurement, namely the years of healthy life lost to illness and death. This permits direct comparison of various diseases and injuries and enables prioritization of preventive and interventional measures. Metrics of disease burden are therefore important indicators of population health that are increasingly being incorporated into national and international health information systems.
The Global Burden of Disease Study (GBD) offers a good overview of the disease burden worldwide (4), but preliminary analyses of the figures for Germany show that the calculations have not always been based on the best data available (5, 6, 7). For this reason, independent burden of disease studies are being carried out in various countries (8, 9, 10). The project “BURDEN 2020: Burden of disease in Germany at the national and regional Level” has the long-term goal of creating a reliable and transparent information source for policy makers based on reproducible data and methods (11).
The aim of the analyses presented here was to calculate the mortality component of disease burden in Germany. In conventional accounts the ranking of the principal causes of death is based on the number of deaths (12, 13). The calculation of years of life lost to death (YLL), however, focuses on the analysis of lost life years rather than the number of deaths.
The consequences of death for population health only become fully visible in terms of YLL, because comparatively rare diseases that lead to death at a young age can result in relatively high numbers of YLL (14, 15, 16). Therefore, the metric YLL can help to answer the question of which diseases should be focused on to minimize the loss of life span and further increase life expectancy. This permits differentiated prioritization of diseases and injuries, particularly with regard to the recognition of age- and sex-specific needs for prevention and care.
The data source for our analyses were the German cause of death statistics for the year 2017 (17), which classify all deaths by sex, age, place of residence (community), and cause of death, with the underlying condition coded according to the International Classification of Diseases and Related Health Problems (ICD-10, WHO, 2016).
Handling of non-informative ICD-10 codes
Overall, 25.8% of the cases in the cause of death statistics for 2017 had “non-informative” ICD-10 codes, i.e., the coding yielded insufficient information regarding the underlying cause of death (13). Non-informative codes may describe sequelae of the underlying illness, symptoms of disease, or unspecific causes of death, or may contain implausible age or sex assignments. Classification into informative and non-informative ICD-10 codes and correction of coding were based on the GBD (4, 18, 19, 20). For cases of death with a non-informative ICD code, assumptions were made regarding the actual causes of death and so-called target codes were defined.
The uncertainty in the estimation of YLL that arose from the redistribution of non-informative ICD-10 codes into informative codes is represented by an uncertainty interval (UI) (eBox 1). Both case numbers and lost life span are therefore reported as ranges. Redistribution and the uncertainty concept are described in detail elsewhere (12).
Figures for deaths and YLL before and after redistributing non-informative ICD-10 codes were compared (eTable 1). For specific causes of death the YLL were classified according to age, sex, and spatial planning region (eBox 2 and eTable 2).
Classification of causes of death
As in the GBD, the ICD-10 codes from the cause of death statistics were classified into cause of death groups.
At the uppermost level, three groups were distinguished:
- A) Communicable, maternal, neonatal, and nutritional diseases (in short: communicable diseases)
- B) Non-communicable diseases
- C) Accidents and Injuries.
At level 2, the non-communicable diseases, for example, are subdivided into neoplasms, cardiovascular diseases, etc.
Further subdivision takes place at level 3. For instance, specific neoplasms are distinguished (lung cancer, breast cancer, and so on) (21) (eTable 3).
A total of 932 272 deaths were registered in Germany in 2017. In 691 467 cases, the ICD-10 codes for cause of death were informative and did not need redistribution (eTable 1). Women made up just over half of the deaths recorded (women: 474 508 deaths, 50.9%; men: 457 761 deaths, 49.1%); three stillbirths were excluded. Group A (communicable diseases) contained on average 4% of cases (36 929; UI: 36 707–37 123). Group B accounted for 91.2% of all deaths (850 534; UI: 850 228–850 875), group C for 4.8% (44 805; UI: 44 585–45 002). The conversion of non-informative ICD-10 codes to informative codes had a particularly pronounced effect on the size of the communicable diseases group, which increased from 10 091 deaths (1.5%) to 36 929 deaths (4.0%) (eTable 1).
Altogether, approximately 11 628 000 years of life were lost to death in 2017: 4 981 000 YLL (42.8%) in women and 6 647 000 YLL (57.2%) in men. The age-specific distribution of the deaths and the YLL shows that proportionally the highest number of deaths was found in the age group 90+ years (19.1%; Figure 1). The highest proportion of YLL was found in the age group 75–79 years (14.3%). Although the absolute number of fatalities rises with increasing age, deaths at a lower age, measured in YLL, exert a greater impact on population health because life expectancy decreases with increasing age. While only 14.7% of deaths occurred in persons under 65 years of age, these cases made up 38.3% of the YLL. In relative numbers (per 100 000 people in the population), the highest burden of disease is in the oldest age group (eFigure 1). In a sensitivity analysis, the YLL were calculated with sex-specific life expectancies (eFigure 2). The number of YLL was then correspondingly lower in men, but developed comparably with increasing age.
At level 2, the disease groups responsible for the highest proportions of total YLL were (malignant) neoplasms (35.2%) and cardiovascular diseases (27.6%). Significant contributions were also made by other groups, foremost digestive diseases (5.8%) and neurological disorders (5.7%). Infectious diseases played a relatively small part. However, the redistribution of non-informative to informative ICD-10 codes resulted in particularly high increases in the case numbers for respiratory infections (+ 1130%), HIV/AIDS and sexually transmitted infections (+ 230%), and diabetes and kidney diseases (+ 80%). In the case of the respiratory infections, the rise was explained mainly by high number of unspecific pulmonary infections (eTable 1) (12).
Looking at the distribution of the YLL at level 2 eTable 1), different disease patterns are evident in different age groups (Figure 2). In 15- to 29-year-olds, the highest proportions of YLL were accounted for by self-harm and violence (27.9–30.4%) and transport injuries (15.5–27.7%). As for (malignant) neoplasms, they played a significant part in children, less so in the 15- to 29-year-olds (11.1–18.4%), but their contribution grew markedly with increasing age. In the age group 60–64 years, almost half (48.2%) of the YLL were due to (malignant) neoplasms. In the uppermost age group (90+ years), in contrast, cardiovascular diseases were to blame for 49.8% of the YLL. The second place in this age category (12.4%) was occupied by the group of neurological disorders, including, for example, dementia.
At level 3, ischemic heart disease was the most important cause of death for both women and men in terms of both number of deaths and YLL (top 20 in Figure 3). Places 2 and 3 were occupied by lung cancer and stroke, respectively, for YLL and by stroke and Alzheimer’s disease with other dementias for number of deaths. The two rankings differed considerably from one another, particularly in women (eFigure 3, eTable 4). Breast cancer was ranked second for YLL and fifth for number of deaths, followed in third place for YLL by lung cancer, which ranked seventh for number of deaths. Bowel cancer (place 7 versus 9) and pancreatic cancer (place 8 versus 14) were further examples of malignancies that ranked higher for YLL than for number of deaths. On the other hand, cardiovascular diseases such as stroke (place 4 versus 2), and hypertensive heart disease (place 9 versus 4), which tend to lead to death later in the life span, ranked lower for YLL than for number of deaths. This was also true for Alzheimer’s disease and other dementias (place 5 versus 3).
For men, the four highest-ranking diagnoses were identical for YLL and number of deaths (ischemic heart disease, lung cancer, stroke, and chronic obstructive pulmonary disease [COPD]). Some other diagnoses, however, ranked much higher for YLL than for number of deaths: chronic liver disease (place 5 versus 8), self-harm (place 6 versus 14), pancreatic cancer (place 9 versus 11), alcohol use disorders (place 10 versus 20). Diseases such as prostate cancer, Alzheimer’s disease and other dementias, diabetes mellitus, and lower respiratory infections (particularly pulmonary infections) ranked lower for YLL than for number of deaths.
The five causes of death accounting for the most YLL varied considerably depending on the age group. While the low numbers of deaths in childhood and adolescence were dominated by congenital and neonatal disorders and trauma, age-associated diseases became the major causes from middle age onwards (Figure 4). Especially men showed high YLL figures for road injuries and self-harm in the period from childhood to early adulthood (eFigure 5). Although the numbers of deaths were low, each fatality caused a particularly high number of YLL. From middle age onwards, elevated burdens of disease were caused by breast cancer in women (eFigure 6) and by lung cancer and chronic liver disease (above all cirrhosis). Starting in the fifth decade of life, ischemic heart disease, COPD, lung cancer, and stomach cancer caused high YLL. At advanced age stroke and dementias became increasingly important for both sexes, accompanied in women over 80 years of age by chronic kidney failure and hypertensive heart disease.
The region with the highest absolute number of YLL (463 911) in 2017 was Berlin. The highest rate, however, was found in the Anhalt-Bitterfeld-Wittenberg spatial planning region, with 20 528 YLL per 100 000 inhabitants. Munich had the lowest YLL rate, with 10 279 per 100 000 inhabitants. Even after age standardization, clear north–south and east–west gradients could be discerned. The highest YLL figures were calculated for Bremerhaven, Anhalt-Bitterfeld-Wittenberg, and Altmark, the lowest for Stuttgart, Oberland, and Munich (Figure 5).
The declared goal of public health and surveillance is to analyze the health of the population and contribute to improvement by showing where action is needed (22, 23). Causes of death yield important information with regard to the potential for prevention at population level, so detailed knowledge of the patterns of mortality is indispensable. If life expectancy is to increase any further, however, data are needed not only on the number of deaths, but also on how much of the life span is lost to individual diseases.
In the framework of the BURDEN 2020 project, wide-reaching analyses of the YLL in Germany were conducted. One of the central findings is that even with the ongoing aging of the population, a significant proportion of the YLL is accounted for by young and intermediate age bands. While only 14.7% of the deaths occurred in persons under 65 years old, 38.3% of the YLL were in this age group.
Altogether, this leads to the major causes of death being ranked in a different order for YLL than for numbers of deaths. While ischemic heart disease is the most important cause of death in both respects, breast cancer and lung cancer in women rank higher for YLL than for numbers of fatalities. In men, on the other hand, chronic liver disease, alcohol use disorders, and self-harm come to the fore. The findings show the need for the provision of prevention programs early in life, for the benefit of both young and old. Accidents, injuries, and self-harm, together with alcohol-associated causes of death, are responsible for an appreciable burden of disease in the young. The need for early prevention measures is also evident with regard to causes of death typically affecting the elderly, such as stroke, in order to prevent accumulation of risks over the life course and raise both life expectancy and the quality of life at an advanced age.
The analyses presented here do have limitations. Assumptions had to be made regarding the probable cause of death in the case of non-informative ICD-10 codes, introducing uncertainty. For this reason we calculated uncertainty intervals (Figure 3, eBox 1, eTable 1, eTables 4–5, eFigures 3–4). Since the German cause of death statistics contain no further information about the fatality (e.g., multicausal data), no individual correction of non-informative codes was possible on this basis. Furthermore, for all deaths with informative ICD-10 codes it was assumed that the cause was correctly coded, which was not necessarily the case.
Compared with the results of the GBD for Germany, the BURDEN 2020 project showed minor discrepancies with regard to the ranking of causes of death for YLL. The added value of the BURDEN 2020 calculations lies in the use of national life tables. The YLL are calculated on the basis of empirically attainable residual life expectancies, so the data portray a genuine potential for prevention. Additionally, region-by-region evaluation is now possible for all causes of death, in contrast to the GBD.
The differences in results between BURDEN 2020 and the GBD are partly due to their use of different procedures for the redistribution of non-informative codes (12). In addition, the GBD uses modeling, e.g., for extrapolation of data. Another important difference is that BURDEN 2020 uses different life expectancies than the GBD. There are currently various ways of selecting life expectancy (eBox 2), which has a crucial effect on YLL figures (24).
For each age group in every country surveyed, the GBD used the highest life expectancy (eBox 2) found anywhere in the world (4). For example, the residual life expectancy for 70- to 74-year-olds is assumed to be 20.3 years, against the assumption of 15.8 years in our study on the basis of deaths in Germany. Other burden of disease studies in other nations have also given preference to their country-specific life expectancies in order to achieve realistic results based on empirical findings with regard to fatalities (25, 26).
In common with the GBD (4) and other burden of disease studies (27), we used the same life expectancy for both sexes. A similarly high life expectancy was thus viewed as attainable, and men and women were compared directly with one another. Research data indicate that the biological component of the difference in life expectancy is small, probably amounting to less than a year. The rest of the difference in life expectancy is due to differences in health-related behavior and in the use of medical care services (28, 29) (eBox 2). The result is that more YLL are documented for men, who on average die earlier, than for women.
This analysis from the BURDEN 2020 study provides detailed findings regarding the years of life lost as a result of all ICD-coded deaths from disease and injury. The methods used were adjusted to the situation in Germany. This permitted, for the first time, the calculation of YLL at regional level, greatly expanding the utility of future burden of disease analyses (30). We thus now have a new evidence base that will form a cornerstone of future monitoring and is being made available for reference as part of an interactive visualization tool. Furthermore, the indicators have been integrated into the diabetes surveillance and the future non-communicable disease (NCD) surveillance programs of the Robert Koch Institute (31). Time series and prognostic models should now be built onto this method to enable the depiction of long-term trends, the sketching of scenarios for future developments, and to accompany measures to enhance population health.
The study “BURDEN 2020: Burden of disease in Germany at the national and regional level” is supported by the innovation fund of the Federal Joint Committee (project number 01VSF17007).
BURDEN 2020 Study Group: Alexander Rommel, Elena von der Lippe, Annelene Wengler, Michael Porst, Aline Anton, Janko Leddin, Thomas Ziese (Robert Koch Institute), Helmut Schröder, Katrin Schüssel, Gabriela Brückner, Jan Breitkreuz (AOK Scientific Institute), Dietrich Plaß, Heike Gruhl (German Environment Agency)
We thank Ronny Kuhnert for advice on the redistribution of non-informative ICD-10 codes and Martin Thißen for the presentation of YLL at the level of spatial planning regions. We are grateful to the members of our scientific advisory committee for methodological advice.
Conflict of interest statement
The authors declare that no conflict of interest exists.
Manuscript received on 15 September 2020, revised version accepted on 3 February 2021
Translated from the original German by David Roseveare
Dr. rer. Elena von der Lippe
Robert Koch-Institut, Abteilung 2, Epidemiologie und Gesundheitsmonitoring
Nordufer 20, 13353 Berlin, Germany
Cite this as:
Wengler A, Rommel A, Plaß D, Gruhl H, Leddin J, Ziese T, von der Lippe E
on behalf of the BURDEN 2020 Study Group: Years of life lost to death—
a comprehensive analysis of mortality in Germany conducted as part of the BURDEN 2020 project. Dtsch Arztebl Int 2021; 118: 137–44.
eReferences, eTables, eFigures, eBoxes:
Dr. rer. pol. Annelene Wengler, Dr. rer. med. Alexander Rommel, Janko Leddin, Dr. med. Thomas Ziese, Dr. rer. pol. Elena von der Lippe
Department II 1 Environmental Hygiene, German Environment Agency, Berlin: Dr. PH Dietrich Plaß, Heike Gruhl
|1.||Murray CJL, Lopez AD: Measuring global health: motivation and evolution of the Global Burden of Disease Study. Lancet 2017; 390: 1460–4 CrossRef MEDLINE|
|2.||Murray CJL, Ezzati M, Flaxman AD, et al.: GBD 2010: design, definitions, and metrics. Lancet 2012; 380: 2063–6 CrossRef MEDLINE|
|3.||Murray CJL: Quantifying the burden of disease: the technical basis for disability-adjusted life years. Bull World Health Organ 1994; 72: 429–45.|
|4.||Roth GA, Abate D, Abate KH, et al.: Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018; 392: 1736–88 CrossRef MEDLINE PubMed Central|
|5.||Murray CJL, Frenk J, Piot P, Mundel T: GBD 2.0: a continuously updated global resource. Lancet 2013; 382: 9–11 CrossRef MEDLINE|
|6.||Plass D, Vos T, Hornberg C, Scheidt-Nave C, Zeeb H, Krämer A: Trends in disease burden in Germany—results, implications and limitations of the Global Burden of Disease Study. Dtsch Arztebl Int 2014; 111: 629–38 CrossRef MEDLINE PubMed Central|
|7.||Kyu HH, Abate D, Abate KH, et al.: Global, regional, and national disability-adjusted life-years (DALYs) for 359 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018; 392: 1859–922 CrossRef MEDLINE PubMed Central|
|8.||Scotland NH: The Scottish Burden of Disease Study, 2015. Overview report 2017. www.scotpho.org.uk/media/1474/sbod2015-overview-report-july17.pdf (last accessed on 16 February 2021).|
|9.||Marieke Verschuuren, Henk B M Hilderink, Robert A A Vonk, The Dutch Public Health Foresight Study 2018: An example of a comprehensive foresight exercise. Eur J Public Health 2020; 30: 30–5 CrossRef MEDLINE|
|10.||Belgian National Burden of Disease Study (BeBOD): Belgian National Burden of Disease Study (BeBOD). www.sciensano.be/en/projects/belgian-national-burden-disease-study (last accessed on 12 November 2020).|
|11.||Rommel A, von der Lippe E, Plass D, et al.: BURDEN 2020-Burden of disease in Germany at the national and regional level. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2018; 61: 1159–66 CrossRef CrossRef MEDLINE|
|12.||Wengler A, Gruhl H, Plaß D, Leddin J, Rommel A, Lippe Evd: Redistributing ill-defined deaths in the German causes of death statistics. Archives of Public Health 2020; 79 (accepted).|
|13.||Wengler A, Rommel A, Plaß D, et al.: ICD-Codierung von Todesursachen: Herausforderungen bei der Berechnung der Krankheitslast in Deutschland. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2019; 62: 1485–92 CrossRef MEDLINE|
|14.||Plass D, Chau P, Thach T, et al.: Quantifying the burden of disease due to premature mortality in Hong Kong using standard expected years of life lost. BMC Public Health 2013; 13: 863 CrossRef MEDLINE PubMed Central|
|15.||Taksler GB, Rothberg MB: Assessing years of life lost versus number of deaths in the United States, 1995–2015. Am J Public Health 2017; 107: 1653–9 CrossRef MEDLINE PubMed Central|
|16.||Martinez R, Soliz P, Caixeta R, Ordunez P: Reflection on modern methods: years of life lost due to premature mortality—a versatile and comprehensive measure for monitoring non-communicable disease mortality. Int J Epidemiol 2019; 48: 1367–76 CrossRef MEDLINE PubMed Central|
|17.||Statistisches Bundesamt: Todesursachenstatistik. DOI: 10.21242/23211.2017.00.00.1.1.0. 2017.|
|18.||Naghavi M, Makela S, Foreman K, O‘Brien J, Pourmalek F, Lozano R: Algorithms for enhancing public health utility of national causes-of-death data. Popul Health Metr 2010; 8: 9 CrossRef MEDLINE PubMed Central|
|19.||Lozano R, Naghavi M, Foreman K, et al.: Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012; 380: 2095–128 CrossRef MEDLINE|
|20.||Naghavi M: Master Cause List for GBD 2019. http://ghdx.healthdata.org/record/ihme-data/gbd-2019-cause-icd-code-mappings (last accessed on 16 Feburary 2021).|
|21.||Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2017 (GBD 2017): Cause, REI, and location hierarchies. Seattle, United States of America. In: Institute for Health Metrics and Evaluation (IHME). (ed.) 2018. http://ghdx.healthdata.org/record/ihme-data/gbd-2017-cause-rei-and-location-hierarchies (last accessed on 16 February 2021).|
|22.||American Public Health Association: Public health code of ethics. Washington: APHA; 2019. www.apha.org/-/media/files/pdf/membergroups/ethics/code_of_ethics.ashx (last accesses 16 Febraury 2021).|
|23.||World Health Organisation: WHO Guidelines on ethical issues in public health surveillance. Geneva: WHO; 2017. www.who.int/ethics/publications/public-health-surveillance/en/ (last accessed 16 Februaty 2021)|
|24.||Devleesschauwer B, McDonald SA, Speybroeck N, Wyper GMA: Valuing the years of life lost due to COVID-19: the differences and pitfalls. Int J Public Health 2020; 65: 719–20 CrossRef MEDLINE PubMed Central|
|25.||Mesalles-Naranjo O, Grant I, Wyper GMA, et al.: Trends and inequalities in the burden of mortality in Scotland 2000–2015. PLoS One 2018; 13: e0196906 CrossRef MEDLINE PubMed Central|
|26.||National Institute for Public Health and the Environment (RIVM): Integrative Measures for the Public Health Foresight Study (VTV) 2018. www.vtv2018.nl/sites/default/files/2018-11/20181108%20Background%20report%20Integrative%20Meassures%20VTV-2018.pdf (last accessed 16 February 2021).|
|27.||Cornez A, Devleesschauwer B: Belgian national burden of disease study. Guidelines for the calculation of DALYs in Belgium. Brussels: sciensano 2020. www.sciensano.be/en/biblio/belgian-national-burden-disease-study-guidelines-calculation-dalys-belgium (last accessed 16 February 2021).|
|28.||Luy M: Causes of male excess mortality: insights from cloistered populations. Population and Development Review 2003; 29: 647–76 CrossRef|
|29.||Rogers RG, Everett BG, Onge JMS, Krueger PM: Social, behavioral, and biological factors, and sex differences in mortality. Demography 2010; 47: 555–78 CrossRef MEDLINE PubMed Central|
|30.||Robert Koch-Institut: BURDEN 2020: Potenzial und Nutzen. www.rki.de/DE/Content/Gesundheitsmonitoring/Studien/Krankheitslast/Potenzial/burden_potenzial_node.html (last accessed on 2 September 2020).|
|31.||Robert Koch-Institut: Diabetes Surveillance: Verlorene Lebensjahre (YLL). www.diabsurv.rki.de/Webs/Diabsurv/DE/indikatoren/4-37_Verlorene_Lebensjahre_YLL.html?nn=11418894 (last accessed on 2 September 2020).|
|e1.||Statistisches Bundesamt: Sterbetafeln 2016/2018, nach Bundesländern, Durchschnittliche Lebenserwartung (Periodensterbetafel). www.genesis.destatis.de/genesis/online (last accessed on 4 May 2020).|
|e2.||Rau R, Schmertmann CP: District-level life expectancy in Germany. Dtsch Arztebl Int 2020; 117: 493–9 VOLLTEXT|
|e3.||Bundesinstitut für Bau-, Stadt- und Raumforschung: Indikatoren und Karten zur Raum- und Stadtentwicklung (INKAR) Datenbank. www.inkar.de/2020 . (last accessed on 18 February 2021).|
Deutsches Aerzteblatt Online, 202110.3238/arztebl.m2021.0258
Deutsches Aerzteblatt Online, 202110.3238/arztebl.m2021.0152
Deutsches Aerzteblatt Online, 202110.3238/arztebl.m2021.0239
Deutsches Aerzteblatt Online, 202110.3238/arztebl.m2021.0244