Parkinson's disease is a long-term degenerative disorder of the central nervous system that involves a reduction in the dopamine production in the compact part of the substantia nigra in the brain. It includes among its symptoms tremor, rigidity, bradykinesia, alteration of posture, loss of balance, akinesia and freezing, but also dementia, depression, anxiety and drowsiness. The evaluation of the severity of the pathology takes place through particular clinical scales: the most used are UPDRS and H&Y. The timed up and go test (TUG) is a clinical test in which the subject, that initially sits on a chair, rises, walks three meters, turns around an obstacle, comes back and sits down again. Since its first description, and especially after the introduction of wearable inertial systems (IMU), the TUG has shown great reliability and versatility in the study of pathologies that affect the motor system and dozens of works in literature use it with the most varied purposes. Studies that treat Parkinson's disease in particular have made massive use of the instrumented TUG (iTUG) to quantify patients' risk of falling, to assess the effectiveness of treatments, to diagnose the disease in the earliest stages when symptoms are low prominent and to evaluate correlations between the performance of the iTUG and other elements of the pathology. The spread of IMUs has led also to the introduction of an instrumented walking (iwalking) test as a low-cost and easy to implement alternative to the traditional 3D gait analysis with optoelectronic systems. In the last decade in particular this test has been widely used to study Parkinson's disease. In literature there are several attempts to establish reference values for the performance of these tests, but ranges for all the parameters derived from the tests and values relating to pathological subjects have not yet been calculated. The aim of this thesis work is to establish reference values for the characteristic parameters of a iTUG and walking test for subjects affected by Parkinson's disease. Reference values allow a clinician to evaluate a patient’s performance by comparing it with the normality of the population he belongs to. The data used for this study come from the movement analysis laboratory of the University of Cagliari and from the San Raffaele Clinic of Cassino (FR). 41 pathological subjects from Cassino and 22 pathological subjects from Cagliari performed in total 184 iTUG trials. 31 pathological subjects from Cassino performed in total 122 iwalking trials. 142 healthy subjects from Cagliari performed 142 iTUG trials. The pursuit of the reference values was carried out by defining, at first, the appropriate classes of sex, age and progression of the pathology for which specific ranges can be identified. The classes assumed for the iTUG tests are men, women, age between 50 and 59, age between 60 and 69, age between 70 and 79, H&Y score ≤ 2.5, H&Y = 3, H&Y = 4, UPDRS III between 0 and 27, UPDRS III between 28 and 54 and UPDRS III between 55 and 81; those hypothesized for iwalking tests are men, women, ages 50 to 59, age between 60 and 69, age between 70 and 79, H&Y = 3, H&Y = 4, UPDRS III between 28 and 54 and UPDRS III between 55 and 81. Ages above 80 and below 50 were not considered due to the reduced number of data available; for the same reason H&Y scores of 1, 2 and 2.5 have been merged into one class. For each parameter of the two tests, statistically significant differences between the classes hypothesized were checked in order to define those that can be adopted for the construction of the ranges. A Shapiro-Wilk normality test was initially carried out on each possible class; the non-Gaussian distribution of most classes led to the adoption of non-parametric methods. In particular, the Mann-Whitney test was used: it’s a non-parametric test which sets as null hypothesis the belonging of the two samples to the same population, rejecting it for a p-value < 0.05. Mann-Whitney tests were carried out between the groups of values from pathological iTUG trials for men and women, H&Y ≤ 2.5 and H&Y = 3, H&Y = 3 and H&Y = 4, UPDRS III between 0 and 27 and UPDRS III between 28 and 54, UPDRS III between 28 and 54 and UPDRS III between 55 and 81, ages between 50 and 59 and ages between 60 and 69, ages between 60 and 69 and ages between 70 and 79. Mann-Whitney tests were carried out between the groups of values from healthy iTUG trials for men and women and ages between 60 and 69 and ages between 70 and 79. Mann-Whitney tests were carried out between the groups of values from iwalking trials for men and women, H&Y = 3 and H&Y = 4, UPDRS III between 28 and 54 and UPDRS III between 55 and 81, age between 50 and 59 and age between 60 and 69, age between 60 and 69 and age between 70 and 79. The data considered intrinsically present a great pathological variability; for this reason it was decided to keep all the data in representation of this variability. Only a patient who systematically deviated sharply from the other subjects was treated as an outlier. Reference values for the overall population and for the classes adopted were calculated for each parameter of the two tests. The ranges was constructed using the main non-parametric method for defining reference values that involves the assumption of the 2.5th and 97.5th percentile as lower and upper limits. The results were presented in summary tables accompanied by mean and standard deviation of each adopted class. The ranges produced appear much wider than those found in literature, to confirm the great intrinsic variability of the pathological data analyzed. For the TUG it is observed that the use of the UPDRS III scale allows the TUG test parameters to more fully describe the course of Parkinson's disease. There are significant differences related to gender. Women take more time both in the sit to walk phase, with 1.96 s (0.82) against 1.66 s (0.76), and in the turn to sit phase, with 1.93 s (0.94) against 1.72 s (0.86). Men present higher accelerations during the sit to walk phase in the vertical plane, with 2.26 m/s2 (1.57) against 1.64 m/s2 (1.37) and in the mediolateral plane, with 0.91 m/s2 (0, 74) against 0.80 m/s2 (0.69), and during the turn to sit phase in the vertical plane, with 4.69 m/s2 (3.01) against 3.58 m/s2 (2.36). At the same time, men have, during the sit to walk phase, a lower width of extension, with 7.96° (7.72) against 15.25° (8.32), and flexion, with 33.83° (13.22) against 40.10° (11.36) and, during the turn to sit phase, a lower width of flexion, with 9.44° (9.23) against 12.81° (9.79). The comparison with the healthy subjects excludes the possibility that this can be traced back to the physiological difference between the sexes. Gender differences can therefore be attributed to the differences in the incidence and course of Parkinson's disease that many studies in literature recognize between the two sexes. Similar observations can be made regarding age, for which the same comparison was made: the results emerged in the analysis of pathological data can be attributed to the effect of Parkinson's disease and are not attributable to physiological conditions linked to aging. For the instrumented walking it is observed, even if the analysis is conducted only on a limited number of classes related to the stage of the desease, a substantial interchangeability of the two clinical scales. Gender differences are not as noticeable as what was observed for the TUG, and however only few parameters are involved. Finally, no parameter was globally effective in describing the aging of the subject. Possible future developments could include a collection of a higher number of data that can guarantee a complete analysis covering all ages and stages of the pathology. The observations made on gender differences could be extended by investigating mixed classes divided by age and gender or stage of the pathology and gender, or by comparing the differences found in TUG and walking parameters between the sexes with the effect that in literature is attributed to the disease on the muscles and mechanisms involved in that specific parameter. Finally, similar studies could be conducted on specific populations in response to particular clinical needs.
La malattia di Parkinson è una patologia cronica neurodegenerativa che vede una riduzione della produzione di dopamina nella parte compatta della sostanza nera del cervello. È una malattia che annovera tra i suoi sintomi tremore, rigidità, bradicinesia, alterazione della postura, perdita di equilibrio, acinesia e freezing, ma anche demenza, depressione, ansia e sonnolenza. La valutazione dello stato e della gravità della patologia si avvale di particolari scale cliniche, principalmente UPDRS e H&Y. Il timed up and go test (TUG) è un test clinico in cui il soggetto inizialmente seduto si alza, cammina per tre metri, gira intorno ad un ostacolo, torna indietro e si siede nuovamente. Fin dalla sua prima descrizione, e soprattutto con l’introduzione della possibilità di strumentare il test con sistemi inerziali indossabili (IMU), il TUG ha mostrato grande affidabilità e versatilità nello studio di patologie che inficiano le funzioni motorie ed è utilizzato in decine di lavori presenti in letteratura con le più svariate finalità e modalità. Studi che trattano la malattia di Parkinson in particolare hanno fatto un uso massiccio del TUG strumentato (iTUG) per quantificare il rischio di caduta dei pazienti, valutare l’efficacia di un trattamento e la progressione della malattia, diagnosticare la patologia fin dalle prime fasi quando i sintomi sono poco evidenti e valutare correlazioni tra la performance dell’iTUG e altri elementi della patologia. La diffusione degli IMU nella pratica clinica ha portato anche all’introduzione di una prova di cammino strumentata (iwalking) come alternativa low cost e attuabile ovunque alla tradizionale gait analysis 3D con sistemi optoelettronici. L’ultimo decennio in particolare ha visto crescere l’applicazione di questo test allo studio della malattia di Parkinson In letteratura sono presenti diversi tentativi di stabilire valori di riferimento per le performance di questi test, ma intervalli completi per tutti i parametri ricavati dalle prove e valori relativi a soggetti patologici non sono ancora stati calcolati. L’obiettivo di questo lavoro di tesi è costruire intervalli di riferimento per i parametri caratteristici di una prova di iTUG e iwalking per soggetti affetti da malattia di Parkinson. I valori di riferimento rispondono alla necessità in ambito clinico di valutare se la performance di un paziente si discosti o meno dalla normalità della popolazione di appartenenza. I dati utilizzati per questo lavoro di tesi provengono dal laboratorio di analisi del movimento dell’Università degli Studi di Cagliari e dalla Casa di Cura San Raffaele di Cassino (FR). 41 soggetti patologici del centro di Cassino e 22 soggetti di Cagliari hanno effettuato un totale di 184 prove di iTUG registrate. 31 soggetti patologici del centro di Cassino hanno effettuato un totale di 122 prove di iwalking. 142 soggetti sani del centro di Cagliari, infine, hanno effettuato un totale di 142 prove di iTUG. La ricerca degli intervalli di riferimento è stata condotta definendo, in un primo momento, le opportune classi di ripartizione per cui trovare i valori considerando le variabili sesso, età e avanzamento della patologia. Le classi ipotizzate per le prove di iTUG sono uomini, donne, età tra 50 e 59 anni, età tra 60 e 69 anni, età tra 70 e 79 anni, punteggio H&Y ≤ 2.5, H&Y = 3, H&Y = 4, UPDRS III tra 0 e 27, UPDRS III tra 28 e 54 e UPDRS III tra 55 e 81; quelle ipotizzate per le prove di iwalking sono uomini, donne, età tra 50 e 59, età tra 60 e 69 anni, età tra 70 e 79 anni, H&Y = 3, H&Y = 4, UPDRS III tra 28 e 54 e UPDRS III tra 55 e 81. Età superiori a 80 e inferiori a 50 non sono state considerate per via della ridotta numerosità dei dati a disposizione; per la stessa ragione punteggi di H&Y pari a 1, 2 e 2.5 sono stati accorpati in un’unica classe per le prove di iTUG e non sono stati considerati per le prove di iwalking. Per ogni parametro delle due prove si sono cercate differenze statisticamente rilevanti tra le classi ipotizzate per definire quelle adottabili per la costruzione degli intervalli. Su ogni possibile classe è stato inizialmente effettuato un test di normalità di Shapiro-Wilk; la distribuzione non gaussiana e la numerosità ridotta di diverse classi hanno portato all’adozione di metodi non parametrici. In particolare si è utilizzato il test di Mann-Whitney che pone come ipotesi nulla l’appartenenza dei due campioni alla medesima popolazione rifiutandola per un p-value < 0.05. Per le prove di iTUG dei soggetti patologici si sono effettuati test tra i gruppi di valori relativi a uomini e donne, H&Y ≤ 2.5 e H&Y = 3, H&Y = 3 e H&Y = 4, UPDRS III tra 0 e 27 e UPDRS III tra 28 e 54, UPDRS III tra 28 e 54 e UPDRS III tra 55 e 81, età tra 50 e 59 e età tra 60 e 69, età tra 60 e 69 e età tra 70 e 79. Per le prove di iTUG dei soggetti sani si sono effettuati test tra i gruppi di valori relativi a uomini e donne, età tra 60 e 69 e età tra 70 e 79. Per le prove di iwalking si sono effettuati test tra i gruppi di valori relativi a uomini e donne, H&Y = 3 e H&Y = 4, UPDRS III tra 28 e 54 e UPDRS III tra 55 e 81, età tra 50 e 59 e età tra 60 e 69, età tra 60 e 69 e età tra 70 e 79. Data la particolarità dei dati considerati, che presentano intrinsecamente una grande variabilità, si è deciso di mantenere tutti i dati in rappresentanza di questa variabilità trattando da outlier solamente un paziente che sistematicamente si discostava nettamente dagli altri soggetti. Per ogni parametro delle due prove si sono calcolati gli intervalli di riferimento per la popolazione complessiva e per le classi adottate. L’intervallo è stato costruito utilizzando il principale metodo non parametrico per la definizione di intervalli di riferimento che prevede l’assunzione del 2.5esimo e 97.5esimo percentile a limiti inferiore e superiore. I risultati sono stati presentati in tabelle riassuntive accompagnati da media e deviazione standard di ogni classe adottata. Gli intervalli prodotti appaiono molto più ampi di quelli presenti in letteratura a conferma della grande variabilità intrinseca dei dati patologici analizzati. Per l’iTUG si osserva che, per quanto riguarda le classi legate all’avanzamento della patologia, l’utilizzo della scala UPDRS III consente ai parametri della prova di iTUG di descrivere in maniera più completa il decorso della malattia di Parkinson. Emergono inoltre notevoli differenze legate al genere. Le donne impiegano più tempo sia in fase di alzata, con 1.96 s (0.82) contro 1.66 s (0.76), che in fase di seduta, con 1.93 s (0.94) contro 1.72 s (0.86). Gli uomini presentano accelerazioni maggiori in alzata sul piano verticale, con 2.26 m/s2 (1.57) contro 1.64 m/s2 (1.37) e sul piano mediolaterale, con 0.91 m/s2 (0.74) contro 0.80 m/s2 (0.69), e in seduta sul piano verticale, con 4.69 m/s2 (3.01) contro 3.58 m/s2 (2.36). Allo stesso tempo gli uomini registrano un’ampiezza inferiore di estensione, con 7.96° (7.72) contro 15.25° (8.32), e flessione, con 33.83° (13.22) contro 40.10° (11.36) in alzata e di flessione in seduta, con 9.44° (9.23) contro 12.81° (9.79). Il confronto con il campione di soggetti sani esclude l’eventualità che questo possa essere ricondotto alla fisiologica diversità tra i sessi. Le differenze di genere emerse possono essere quindi attribuite alle differenze nell’incidenza e nel decorso della malattia di Parkinson che numerosi studi in letteratura riconoscono tra i due sessi. Osservazioni analoghe possono essere compiute riguardo l’età, per la quale è stato condotto lo stesso confronto: i risultati emersi nell’analisi dei dati patologici possono essere attribuiti all’effetto della malattia di Parkinson e non sono riconducibili a condizioni fisiologiche legate all’invecchiamento. Per l’instrumented walking si osserva invece, seppur con un’analisi condotta solamente su un numero limitato di classi che non descrive la completa evoluzione della malattia, una sostanziale interscambiabilità delle due scale cliniche. Le differenze di genere rappresentano un fenomeno ridotto e limitato solamente a pochi parametri rispetto a quanto osservato per l’iTUG. Nessun parametro, infine, è risultato globalmente efficace nel descrivere l’invecchiamento del soggetto. Possibili sviluppi futuri potrebbero prevedere una raccolta di un numero superiore di dati che possano garantire un’analisi completa che copra tutte le età e tutte le fasi della patologia. Potrebbero poi essere approfondite le osservazioni compiute sulle differenze di genere indagando classi di ripartizioni miste suddivise per età e genere o avanzamento e genere oppure confrontando le differenze riscontrate tra i sessi nei diversi parametri con l’effetto che in letteratura è riconosciuto alla malattia sui muscoli e i meccanismi coinvolti in quello specifico parametro. Si potrebbero infine condurre studi analoghi su popolazioni specifiche in risposta a esigenze cliniche particolari.
Valori di riferimento per la malattia di Parkinson relativi a test funzionali strumentati
LUCREZI, FEDERICO
2018/2019
Abstract
Parkinson's disease is a long-term degenerative disorder of the central nervous system that involves a reduction in the dopamine production in the compact part of the substantia nigra in the brain. It includes among its symptoms tremor, rigidity, bradykinesia, alteration of posture, loss of balance, akinesia and freezing, but also dementia, depression, anxiety and drowsiness. The evaluation of the severity of the pathology takes place through particular clinical scales: the most used are UPDRS and H&Y. The timed up and go test (TUG) is a clinical test in which the subject, that initially sits on a chair, rises, walks three meters, turns around an obstacle, comes back and sits down again. Since its first description, and especially after the introduction of wearable inertial systems (IMU), the TUG has shown great reliability and versatility in the study of pathologies that affect the motor system and dozens of works in literature use it with the most varied purposes. Studies that treat Parkinson's disease in particular have made massive use of the instrumented TUG (iTUG) to quantify patients' risk of falling, to assess the effectiveness of treatments, to diagnose the disease in the earliest stages when symptoms are low prominent and to evaluate correlations between the performance of the iTUG and other elements of the pathology. The spread of IMUs has led also to the introduction of an instrumented walking (iwalking) test as a low-cost and easy to implement alternative to the traditional 3D gait analysis with optoelectronic systems. In the last decade in particular this test has been widely used to study Parkinson's disease. In literature there are several attempts to establish reference values for the performance of these tests, but ranges for all the parameters derived from the tests and values relating to pathological subjects have not yet been calculated. The aim of this thesis work is to establish reference values for the characteristic parameters of a iTUG and walking test for subjects affected by Parkinson's disease. Reference values allow a clinician to evaluate a patient’s performance by comparing it with the normality of the population he belongs to. The data used for this study come from the movement analysis laboratory of the University of Cagliari and from the San Raffaele Clinic of Cassino (FR). 41 pathological subjects from Cassino and 22 pathological subjects from Cagliari performed in total 184 iTUG trials. 31 pathological subjects from Cassino performed in total 122 iwalking trials. 142 healthy subjects from Cagliari performed 142 iTUG trials. The pursuit of the reference values was carried out by defining, at first, the appropriate classes of sex, age and progression of the pathology for which specific ranges can be identified. The classes assumed for the iTUG tests are men, women, age between 50 and 59, age between 60 and 69, age between 70 and 79, H&Y score ≤ 2.5, H&Y = 3, H&Y = 4, UPDRS III between 0 and 27, UPDRS III between 28 and 54 and UPDRS III between 55 and 81; those hypothesized for iwalking tests are men, women, ages 50 to 59, age between 60 and 69, age between 70 and 79, H&Y = 3, H&Y = 4, UPDRS III between 28 and 54 and UPDRS III between 55 and 81. Ages above 80 and below 50 were not considered due to the reduced number of data available; for the same reason H&Y scores of 1, 2 and 2.5 have been merged into one class. For each parameter of the two tests, statistically significant differences between the classes hypothesized were checked in order to define those that can be adopted for the construction of the ranges. A Shapiro-Wilk normality test was initially carried out on each possible class; the non-Gaussian distribution of most classes led to the adoption of non-parametric methods. In particular, the Mann-Whitney test was used: it’s a non-parametric test which sets as null hypothesis the belonging of the two samples to the same population, rejecting it for a p-value < 0.05. Mann-Whitney tests were carried out between the groups of values from pathological iTUG trials for men and women, H&Y ≤ 2.5 and H&Y = 3, H&Y = 3 and H&Y = 4, UPDRS III between 0 and 27 and UPDRS III between 28 and 54, UPDRS III between 28 and 54 and UPDRS III between 55 and 81, ages between 50 and 59 and ages between 60 and 69, ages between 60 and 69 and ages between 70 and 79. Mann-Whitney tests were carried out between the groups of values from healthy iTUG trials for men and women and ages between 60 and 69 and ages between 70 and 79. Mann-Whitney tests were carried out between the groups of values from iwalking trials for men and women, H&Y = 3 and H&Y = 4, UPDRS III between 28 and 54 and UPDRS III between 55 and 81, age between 50 and 59 and age between 60 and 69, age between 60 and 69 and age between 70 and 79. The data considered intrinsically present a great pathological variability; for this reason it was decided to keep all the data in representation of this variability. Only a patient who systematically deviated sharply from the other subjects was treated as an outlier. Reference values for the overall population and for the classes adopted were calculated for each parameter of the two tests. The ranges was constructed using the main non-parametric method for defining reference values that involves the assumption of the 2.5th and 97.5th percentile as lower and upper limits. The results were presented in summary tables accompanied by mean and standard deviation of each adopted class. The ranges produced appear much wider than those found in literature, to confirm the great intrinsic variability of the pathological data analyzed. For the TUG it is observed that the use of the UPDRS III scale allows the TUG test parameters to more fully describe the course of Parkinson's disease. There are significant differences related to gender. Women take more time both in the sit to walk phase, with 1.96 s (0.82) against 1.66 s (0.76), and in the turn to sit phase, with 1.93 s (0.94) against 1.72 s (0.86). Men present higher accelerations during the sit to walk phase in the vertical plane, with 2.26 m/s2 (1.57) against 1.64 m/s2 (1.37) and in the mediolateral plane, with 0.91 m/s2 (0, 74) against 0.80 m/s2 (0.69), and during the turn to sit phase in the vertical plane, with 4.69 m/s2 (3.01) against 3.58 m/s2 (2.36). At the same time, men have, during the sit to walk phase, a lower width of extension, with 7.96° (7.72) against 15.25° (8.32), and flexion, with 33.83° (13.22) against 40.10° (11.36) and, during the turn to sit phase, a lower width of flexion, with 9.44° (9.23) against 12.81° (9.79). The comparison with the healthy subjects excludes the possibility that this can be traced back to the physiological difference between the sexes. Gender differences can therefore be attributed to the differences in the incidence and course of Parkinson's disease that many studies in literature recognize between the two sexes. Similar observations can be made regarding age, for which the same comparison was made: the results emerged in the analysis of pathological data can be attributed to the effect of Parkinson's disease and are not attributable to physiological conditions linked to aging. For the instrumented walking it is observed, even if the analysis is conducted only on a limited number of classes related to the stage of the desease, a substantial interchangeability of the two clinical scales. Gender differences are not as noticeable as what was observed for the TUG, and however only few parameters are involved. Finally, no parameter was globally effective in describing the aging of the subject. Possible future developments could include a collection of a higher number of data that can guarantee a complete analysis covering all ages and stages of the pathology. The observations made on gender differences could be extended by investigating mixed classes divided by age and gender or stage of the pathology and gender, or by comparing the differences found in TUG and walking parameters between the sexes with the effect that in literature is attributed to the disease on the muscles and mechanisms involved in that specific parameter. Finally, similar studies could be conducted on specific populations in response to particular clinical needs.File | Dimensione | Formato | |
---|---|---|---|
tesi_federico_lucrezi.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Tesi di laurea Federico Lucrezi
Dimensione
2.47 MB
Formato
Adobe PDF
|
2.47 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/147894