Spatial Patterns in Neurodegenerative Disease’s Hospitalizations in

Portugal (2000-2016)

Mariana Oliveira

1,2 a

, Alberto Freitas

1,2 b

, Ana Cláudia Teodoro

3,4 c

and Hernâni Gonçalves

1,2 d

Center for Health Technology and Services Research (CINTESIS@RISE), Faculty of Medicine, University of Porto,

Rua Doutor Plácido da Costa, s/n 4200-450 Porto, Portugal

Department of Community Medicine, Information and Health Decision Sciences (MEDCIDS), Faculty of Medicine,

University of Porto, Rua Doutor Plácido da Costa s/n, 4200-450 Porto, Portugal

Department of Geosciences, Environment and Land Planning, Faculty of Sciences, University of Porto,

Rua do Campo Alegre 687, 4169-007 Porto, Portugal

Earth Sciences Institute (ICT), Pole of the FCUP, University of Porto, 4169-007 Porto, Portugal

Keywords: Neurodegenerative Diseases, Spatial Epidemiology, Hospitalization Rates.

Abstract: Neurodegenerative diseases, usually arising from the death of nervous system cells, are a rising concern in

the worlds’ population increasing life expectancy. More precisely, the Portuguese population, along with that

of other developed countries, is ageing at a fast rate. The understanding of such diseases’ patterns is of utmost

relevance to help manage the burden it represents in the health system. In this retrospective study, we analysed

over 500 thousand hospitalizations with discharges between 2000 and 2016. We computed age-standardized

hospitalization rates for each neurodegenerative disease. The most prevalent disease in our sample was

Dementia with 43.4% of cases, and the least prevalent was Basal with only 0.2% of cases. The spatial analysis

shows that Santarém and Portalegre (neighbour) districts in central Portugal, have the highest rates. The

increase in hospitalization rates over the study period is also clear when looking at the spatio-temporal

analysis. Although limited by the usage of secondary health data, this study represents a background for other

studies on the field of neurodegenerative diseases, presenting with relevant insight into the spatio-temporal

patterns of each and every neurodegenerative disease in Portugal at the moment.

1 INTRODUCTION

Global life expectancy has been increasing, from 66.8

in 2000 to 73.3 years in 2019 (WHO, 2022) , raising

the concern on neurodegenerative diseases, since age

is one of the main contributing factors to developing

these pathologies. Neurodegeneration often occurs

due to the pathological changes in the human brain

arisen from aging (Spencer et al., 2016), which result

in the death of cells either by necrosis or delayed

apoptosis (Brown et al., 2005).

Among these diseases are Alzheimer’s and other

dementias, which contribute to 12% of the burden in

disability-adjusted life years (DALYs), worsened by

the fact that it is the most common form of late onset

https://orcid.org/0000-0001-7313-9749

https://orcid.org/0000-0003-2113-9653

https://orcid.org/0000-0002-8043-6431

https://orcid.org/0000-0001-6558-8966

dementia globally (Rabinovici, 2019). Along with

Parkinson’s disease, they are the most common

neurodegenerative diseases (de Lau & Breteler, 2006;

Jankovic, 2008). Patients with these diseases are more

susceptible to longer lengths of stay and represent a

higher mortality rate, as well as a higher likeliness to

be discharged to long-term care (Timmons et al.,

2015). Other neurodegenerative diseases include

Multiple Sclerosis, Huntington’s, Basal Ganglia and

Pick’s.

Some risk factors to these diseases have been

already discussed. For instance, atop from age as

presented earlier, gender, hereditarianism and

environmental exposure to pollution are all

established risk factors for many of the

neurodegenerative diseases. For instance, in

Oliveira, M., Freitas, A., Teodoro, A. and Gonçalves, H.

Spatial Patterns in Neurodegenerative Disease’s Hospitalizations in Portugal (2000-2016).

DOI: 10.5220/0011990100003473

In Proceedings of the 9th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2023), pages 79-86

ISBN: 978-989-758-649-1; ISSN: 2184-500X

literature, one third of Alzheimer’s patients are

female, while with Parkinson’s the prevalence is

greater in male patients (de Lau & Breteler, 2006). On

the other hand, Huntington’s disease is entirely

acquired through hereditarianism, and some forms of

Alzheimer’s, mainly early-onset, are also transmitted

through genetics (Price et al., 1998).

Furthermore, several studies have showed the

significance of spatially analysing neurodegenerative

disorders, displaying spatial patterns that uncover

new correlations with variables such as latitude and

environmental pollutants (Oliveira et al., 2020).

Mapping such diseases and understanding the

distribution and characteristics of the patients is of

ultimate importance to plan and mitigate for such

disorders.

To the authors’ best knowledge, this is the first

study developed in Portugal which describes the

hospitalizations by neurodegenerative disorders, both

statistically and spatially. With this retrospective

study with data from 2000 to 2016, we hope to help

provide insight on the distribution of these diseases.

2 MATERIALS AND METHODS

2.1 Study Area

This study was conducted with data from Portugal

mainland, located in the southwest of Europe, with an

area of approximately 89.000 km

and an aging

population of around 10,000,000 inhabitants, 21% of

which are 65 years old or older (INE, 2020). Figure 1

shows the distribution of age groups population for

each district of Portugal, using 10 years-old

categories. Population is also very heterogeneously

distributed, as visible in Figure 2.

2.2 Data Acquisition

2.2.1 Health Data

Hospitalizations of patients with at least one

neurodegenerative pathology coded as a diagnosis,

both principal or secondary, were obtained from

secondary hospitalization data, with discharges

between 2000 to 2016. No further years were

included since the provided data frame included only

hospitalizations from 2000 to 2018, but both 2017 and

2018 were incomplete and would thus result in non-

comparable results. Each record contains the year,

patient’s sex, age and residence (coded up to the

parish), entry and leave dates, length of stay and

coded diagnosis.

Figure 1: Population age distribution per groups of 10 years

for each district (2000 to 2021 mean).

Neurodegenerative pathologies codes were

retrieved from literature review (Kim et al., 2019; Lee

et al., 2019; Min et al., 2020; Rochoy et al., 2019; St

Germaine-Smith et al., 2012; Wetmore et al.,

2019),along with counselling with coding experts. All

resulting International Classification of Diseases

(ICD) codes used are listed in Table 1.

GISTAM 2023 - 9th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 2: Population percentage per district.

2.2.2 Population Data

Population data was retrieved from National Institute

of Statistics (INE) Population Census for the years of

2001, 2011 and 2021, for each Portuguese mainland

parish. These data were then linearly interpolated to

obtain an approximate value of population by year,

further used to normalize health data. District and

municipality population was also computed from the

sum of the contained parishes. Only integer values

were used, due to the nature of the population

variable.

2.3 Data Processing

All data pre-processing, descriptive analysis and

mapping was done using Python programming

language.

Table 1: ICD codes used to retrieve hospitalization data.

Dia

nosis ICD 9 ICD 10

Alzheime

331.0 G30

Pick's 331.1

G31

Lewy 331.8

Creutzfeldt 046 A81.0

Huntington 333.4 G10

Parkinson 332 G20

Motor neuron disease 335 G12.2

Multi

le sclerosis 340 G35

Basal ganglia 333.0 G23

Dementia

Vascula

290.4

F01

F02

F03

Uns

ecifie

294

Senile 331.2

Cerebral de

eneration 331.7

Presenile 290.1

Senile with delusion 290.2

Senile with deliriu

290.3

Alcohol induce

291.2

Dru

induce

292.82

Other 331.9

G32

Health data was loaded and diagnosis codes were

extracted and recoded as the disease they referred to.

For example, ICD-9 code 331.0 and ICD-10 code

G30 were both recoded as ‘Alzheimer’. The category

“Other”, as stated in the ICD, represents the other

neurodegenerative diseases not described elsewhere

by the alternative codes. Age was grouped into three

categories: under 20 years old, 20 to 70 years old, and

over 70 years old. These categories were chosen to

adapt to the available population categories, as well

as to accommodate the distinction between paediatric,

early onset and late onset neurodegenerative diseases

(Fatima et al., 2022).

Afterwards, general descriptive statistics was

computed. Then, data was grouped by either district,

municipality or parish to proceed to mapping, and

population in each administrative area, as well as the

percentage of each age group, were used to compute

age-standardized hospitalization rates (ASHR), using

equation 1 (Curado et al., 2007). This rate allows data

to be normalized and therefore enables comparison of

the rates, both across space and time, with a lower

degree of error.

𝐴𝑆𝐻𝑅 =

𝑁



𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛



×%



(1)

In equation 1, N

AgeGroup

represents the number of

hospitalizations in each age group, Population

represents the total population in the area

corresponding to each age group (for example, each

district) and %AgeGroup the percentage of the age

group in the population. The resulting data was

therefore mapped to establish spatio-temporal

patterns.

Spatial Patterns in Neurodegenerative Disease’s Hospitalizations in Portugal (2000-2016)

3 RESULTS

3.1 Descriptive Statistics

A total of 502 245 hospitalizations were included in

this study. As represented in the graphic in Figure 3,

the most prevalent disease was Dementia (43.4%),

followed by Other (22.7%) and Alzheimer (20.2%).

Basal was the least prevalent disease (0.2%).

Figure 3: Number of hospitalizations and respective

percentage per neurodegenerative disease, relative to the

total number of hospitalizations.

Most patients were female (56.3%) and 81.3% of all

patients were over 70 years old, with only 0.4% of the

cases representing paediatric disorders (under 20

years old). Of these, Motor Neuron was by far the

most prevalent neurodegenerative disease with 748

cases (Table 2).

3.2 Spatial Analysis

As expected, mapping the available data provided

frther insight on the spatio-temporal distribution,

otherwise difficult to grasp from algebraic statistical

methods.

Although the descriptive statistics pointed the

Centre as the area with the most prevalence of overall

neurodegenerative diseases, mostly coded as

“Dementia”, the map of the hospitalization rates

(Figure 4) shows that the most incidence is found in

the district of Bragança, located in the north of

Portugal. Furthermore, Porto district, where a main

metropolis of Portugal is located, has lower

hospitalization rates, contrarily to Lisbon, the other

major metropolis of Portugal. Each individual disease

is also mapped in Figure 5.

Figure 4: Age-standardized hospitalization rates by

neurodegenerative diseases per district for the years 2000 to

2016.

Another relevant aspect to be mapped is the temporal

series. In this, the results clearly show that most

districts exhibited a consistent increase throughout

the years, particularly Portalegre in the southeast

(Figure 6).

4 DISCUSSION

To the best of the authors’ knowledge, this is the first

population based spatio-temporal descriptive study of

neurodegenerative pathologies in Portugal. This work

relies on the availability of nationwide hospitalization

data from public hospital for 17 years, thus

representing a significant sample of patients with

neurodegenerative diseases in Portugal. Furthermore,

the availability of each patient’s residence allows for

a spatial analysis not available as standard practice.

GISTAM 2023 - 9th International Conference on Geographical Information Systems Theory, Applications and Management

Table 2: Descriptive statistics of each neurodegenerative disease included in this study, as well as their total. Categorical data

(sex, age group and residence) is described by its absolute frequency and percentage, while continuous data (age and length

of stay, which are both asymmetrically distributed), is described by its median and interquartile range. The total row describes

the total frequency of each disease and their respective percentage. yo: years old.

As expected, elders were the most prevalent age

group, representing over 81% of the overall patients.

Only for Creutzfeldt, Huntington, Motor Neuron,

Multiple Sclerosis and Pick was the age mean lower

than 65 years old. Most patients were women (56%)

and patients were hospitalized for a median of 8 days.

As for the spatial patterns found, the centre showed a

slightly higher incidence of neurodegenerative

diseases although the rates are similar throughout the

country.

Porto didn’t show especially high rates of

hospitalizations, even though it is one of the most

populated cities of Portugal. This may be due to easier

access to health, which may allow for the better

management of the diseases, avoiding

hospitalization. On the opposite side, the higher rate

of hospitalizations in the interior of the country where

isolation and aging population are combined to hinder

health access may be thus explained. Other reason

may involve lower literacy, which is a known risk

factor for developing neurodegenerative diseases, and

highly inequal between the interior and the coast of

the country (Ávila, 2006).

Furthermore, the aging of the population is well

represented in the higher rates of hospitalizations

along the time period of the study.

4.1 Limitations

One of the major limitations in our work is the fact

that the hospitalizations dataset does not include a

unique patient identifier, to allow us to attempt to

identify patients unequivocally. As such, if a patient

with a neurodegenerative disease

has been

hospitalized more than once, it will appear repeated

in our database, without us being able to realize it and

filter it out. For this reason, data here represented may

be oversampling some disease that would naturally

require more hospitalizations than other. Ideally,

datasets should include a unique patient identifier to

advert this issue.

Alzheimer

Basal

Creutzfeldt

Dementia

Huntington

Lewy

Motor Neuron

Multiple Sclerosis

Parkinson

Pick

Other

Total

Sex

Female

62349

(61.4)

659

(55.8)

1209

(34.1)

127694

(58.6)

736

(47.1)

3458

(49.9)

3272

(45.5)

12062

(67.2)

56977

(50.0)

722

(45.1)

13775

(47.4)

282913

(56.3)

Male

39256

(38.6)

521

(44.2)

2332

(65.9)

90087

(41.4)

828

(52.9)

3474

(50.1)

3919

(54.5)

5898

(32.8)

56877

(50.0)

879

(54.9)

15260

(52.6)

219331

(43.7)

Age (yo)

[76, 85]

[61, 79]

[38, 68]

[76, 87]

[47, 70]

[70, 84]

[53, 73]

[33, 54]

[73, 84]

[60, 77]

[66, 83]

[74, 86]

e Grou

< 20 yo

16 (0.0) 39 (3.3)

116

(3.3)

157

(0.1)

29 (1.9)

130

(1.9)

748

(10.4)

403

(2.2)

59 (0.1) 2 (0.1)

187

(0.6)

1886

(0.4)

20-70 yo

9018

(8.9)

511

(43.3)

2649

(74.8)

25241

(11.6)

1169

(74.7)

1656

(23.9)

4087

(56.8)

16775

(93.4)

20052

(17.6)

956

(59.7)

9740

(33.5)

91854

(18.3)

> 70 yo

92571

(91.1)

630

(53.4)

776

(21.9)

192383

(88.3)

366

(23.4)

5146

(74.2)

2356

(32.8)

782

(4.4)

93744

(82.3)

643

(40.2)

19108

(35.8)

408505

(81.3)

Length of

stay (days)

[4, 13]

[4, 15]

[7, 34]

[5, 14]

[4, 16]

[5, 16]

[3, 13]

[2, 9]

[4, 14]

[6, 24]

[5, 17]

[4, 14]

Residence

North

37116

(36.5)

513

(43.5)

973

(27.5)

79156

(36.3)

437

(27.9)

2687

(38.8)

2396

(33.3)

5411

(30.1)

35190

(30.9)

554

(34.6)

22274

(76.7)

186707

(37.2)

Centre

49750

(49.0)

518

(43.9)

1897

(53.6)

114337

(52.5)

882

(56.4)

3642

(52.5)

3621

(50.4)

9487

(52.8)

61947

(54.4)

901

(56.3)

5225

(18.0)

252207

(50.2)

South

14739

(14.5)

149

(12.6)

671

(18.9)

24288

(11.2)

245

(15.7)

603

(8.7)

1174

(16.3)

3062

(17.0)

16718

(14.7)

146

(9.1)

1536

(5.3)

63331

(12.6)

Total

101605

(20.2)

1180

(0.2)

3541

(0.7)

217781

(43.4)

1564

(0.3)

6932

(1.4)

7191

(1.4)

17960

(3.6)

113855

(22.7)

1601

(0.3)

29035

(5.8)

502245

(100)

Spatial Patterns in Neurodegenerative Disease’s Hospitalizations in Portugal (2000-2016)

Figure 5: Neurodegenerative diseases’ spatial distribution per district.

Reusing data is not always as accurate as

desirable. Nevertheless, it is the best available option

when studying a large population throughout a large

time frame, without the implications of collecting

data prospectively.

4.2 Implications and Future Work

This study may be used as a base for further studies

concerning neurodegenerative pathologies, as it gives

insight on the spatial patterns expected, as well as to

anticipating the statistical distribution of Portuguese

public hospitalization data.

GISTAM 2023 - 9th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 6: Spatio-temporal age-standardized hospitalization rates by neurodegenerative diseases per district for each year, from

2000 to 2016.

5 CONCLUSION

This work describes the spatio-temporal patterns of

age-standardized hospitalization rates of

neurodegenerative diseases, in Portugal mainland,

from 2000 to 2016. The most prevalent disease was

Dementia, and the least prevalent was Basal.

The

aging Portuguese population is notorious on

the temporal analysis, where it becomes clear that

neurodegenerative diseases incidence is increasing.

Also, disparities in health access become clear as

interior districts have higher hospitalization rates than

littoral ones and, mostly, than the larger metropolis.

With this work, we hope that the established

patterns may be of use when working with similar data

sources.

Spatial Patterns in Neurodegenerative Disease’s Hospitalizations in Portugal (2000-2016)

ACKNOWLEDGEMENTS

This research was funded by FCT – Fundação para a

Ciência e Tecnologia, grant number SFRH/BD/

147324/2019 and national MCTES funds and

supported also by National Funds through FCT –

Fundação para a Ciência e a Tecnologia, I.P., within

CINTESIS, R&D Unit (reference UIDB/4255/2020).

Hospitalization data was provided by the Central

Administration of the Health System, I.P. (ACSS).

REFERENCES

Ávila, P. D. (2006). A literacia dos adultos: competências-

chave na sociedade do conhecimento.

Brown, R. C., Lockwood, A. H., & Sonawane, B. R. (2005).

Neurodegenerative diseases: an overview of

environmental risk factors. Environ Health Perspect,

113(9), 1250-1256. https://doi.org/10.1289/ehp.7567

Carvalho, R., Lobo, M., Oliveira, M., Oliveira, A. R.,

Lopes, F., Souza, J., Ramalho, A., Viana, J., Alonso, V.,

Caballero, I., Santos, J. V., & Freitas, A. (2021).

Analysis of root causes of problems affecting the quality

of hospital administrative data: A systematic review and

Ishikawa diagram. Int J Med Inform, 156, 104584.

https://doi.org/10.1016/j.ijmedinf.2021.104584

Curado, M.-P., Edwards, B., Shin, H. R., Storm, H., Ferlay,

J., Heanue, M., & Boyle, P. (2007). Cancer incidence in

five continents, Volume IX. IARC Press, International

Agency for Research on Cancer.

de Lau, L. M., & Breteler, M. M. (2006). Epidemiology of

Parkinson's disease. Lancet Neurol, 5(6), 525-535.

https://doi.org/10.1016/s1474-4422(06)70471-9

Fatima, K., Mehendale, A. M., & Reddy, H. (2022). Young-

Onset Dementia and Neurodegenerative Disorders of

the Young with an Emphasis on Clinical

Manifestations. Cureus, 14(10).

INE, N. S. I. (2020). Projections of resident population in

Portugal. Retrieved 25/01/2023 from https://www.

ine.pt/xportal/xmain?xpid=INE&xpgid=ine_destaques

&DESTAQUESdest_boui=406534255&DESTAQUE

Smodo=2&xlang=en

Jankovic, J. (2008). Parkinson's disease: clinical features

and diagnosis. J Neurol Neurosurg Psychiatry, 79(4),

368-376. https://doi.org/10.1136/jnnp.2007.131045

Kim, S. Y., Min, C., Oh, D. J., & Choi, H. G. (2019). Risk

of neurodegenerative dementia in asthma patients: a

nested case–control study using a national sample

cohort. BMJ Open, 9(10), e030227. https://doi.org/

10.1136/bmjopen-2019-030227

Lee, S. Y., Lim, J. S., Oh, D. J., Kong, I. G., & Choi, H. G.

(2019). Increased risk of neurodegenerative dementia in

women with migraines: A nested case-control study

using a national sample cohort. Medicine (Baltimore),

98(7), e14467. https://doi.org/10.1097/md.00000000000

14467

Min, C., Bang, W. J., Kim, M., Oh, D. J., & Choi, H. G.

(2020). Rheumatoid arthritis and neurodegenerative

dementia: a nested case-control study and a follow-up

study using a national sample cohort. Clinical

Rheumatology, 39(1), 159-166. https://doi.org/10.1007

/s10067-019-04769-x

Oliveira, M., Padrão, A., Ramalho, A., Lobo, M., Teodoro,

A. C., Gonçalves, H., & Freitas, A. (2020). Geospatial

Analysis of Environmental Atmospheric Risk Factors in

Neurodegenerative Diseases: A Systematic Review.

International Journal of Environmental Research and

Public Health, 17(22), 8414. https://www

.mdpi.com/1660-4601/17/22/8414

Price, D. L., Sisodia, S. S., & Borchelt, D. R. (1998). Genetic

Neurodegenerative Diseases: The Human Illness and

Transgenic Models. Science, 282(5391), 1079-1083.

https://doi.org/doi:10.1126/science.282.5391.1079

Rabinovici, G. D. (2019). Late-onset Alzheimer Disease.

Continuum (Minneap Minn), 25(1), 14-33. https://

doi.org/10.1212/con.0000000000000700

Rochoy, M., Bordet, R., Gautier, S., & Chazard, E. (2019).

Factors associated with the onset of Alzheimer's

disease: Data mining in the French nationwide

discharge summary database between 2008 and 2014.

PLoS One, 14(7), e0220174. https://doi.org/10.1371/

journal.pone.0220174

Santurtún, A., Delgado-Alvarado, M., Villar, A., &

Riancho, J. (2016). Patrón geográfico de la mortalidad

por enfermedad de Parkinson en España y su asociación

con los niveles de plomo en el aire [10.1016/

j.medcli.2016.07.022]. Medicina Clínica, 147(11), 481-

487. https://doi.org/10.1016/j.medcli.2016. 07.022

Spencer, P. S., Palmer, V. S., & Kisby, G. E. (2016). Seeking

environmental causes of neurodegenerative disease and

envisioning primary prevention. Neurotoxicology, 56,

269-283. https://doi.org/10.1016/j.neuro.2016.03.017

St Germaine-Smith, C., Metcalfe, A., Pringsheim, T.,

Roberts, J. I., Beck, C. A., Hemmelgarn, B. R.,

McChesney, J., Quan, H., & Jette, N. (2012).

Recommendations for optimal ICD codes to study

neurologic conditions: a systematic review. Neurology,

79(10), 1049-1055. https://doi.org/10.1212/WNL.0b013

e3182684707

Timmons, S., Manning, E., Barrett, A., Brady, N. M.,

Browne, V., O’Shea, E., Molloy, D. W., O'Regan, N. A.,

Trawley, S., Cahill, S., O'Sullivan, K., Woods, N.,

Meagher, D., Ni Chorcorain, A. M., & Linehan, J. G.

(2015). Dementia in older people admitted to hospital:

a regional multi-hospital observational study of

prevalence, associations and case recognition. Age and

Ageing, 44(6), 993-999. https://doi.org/10.1093/

ageing/afv131

Wetmore, J. B., Li, S., Yan, H., Irfan, M., Rashid, N., Peng,

Y., Gilbertson, D. T., & Shim, A. (2019). Increases in

institutionalization, healthcare resource utilization, and

mortality risk associated with Parkinson disease

psychosis: Retrospective cohort study. Parkinsonism

Relat Disord, 68, 95-101. https://doi.org/10.1016/

j.parkreldis.2019.10.018

WHO, W. H. O. (2022). World Health Statistics 2022.

GISTAM 2023 - 9th International Conference on Geographical Information Systems Theory, Applications and Management