This thesis aims to study the effects of isolation caused by the first wave of Covid-19 on the mental health of the elderly population. In particular, the phenomena of depression and loneliness will be studied. This study was based on a panel of data collected through the SHARE survey (Survey of Health, Ageing and Retirement in Europe), and mainly on those related to SHARE-COVID 19. The survey was conducted in 28 different countries, all belonging to the European Union, except Switzerland and Israel. The SHARE survey was created to follow the ageing process of the population; therefore, the target population of the study is composed of people aged 50 and over. The questions of the questionnaire cover different fields, such as physical and mental health, economic situation and social networks of the interviewed individuals. In particular, in the SHARE-COVID 19 survey, the situation before and after the pandemic was analysed. To study the phenomenon of interest, two response variables were created, one for depression and one for loneliness, which detect the worsening (or onset) of these due to the pandemic. These are two dichotomous variables, which were analysed separately; for depression and loneliness respectively, the distinction between the two classes concerns those who state that they feel more depressed and lonelier than in the pre-pandemic period and those who do not state that they have experienced this worsening. Starting from the questions in the questionnaire, we intend to understand as far as possible which are the most significant factors that have influenced these symptoms in the elderly. For example, did the use of social media help respondents to stay in touch with their family? Did it have a positive influence on feeling less lonely? In addition to answering these questions, we tried to outline and discover potential differences between the countries according to the respective restrictions adopted. The first step was to contextualise the study phenomenon, and therefore to analyse the pandemic situation and its evolution in terms of infections and deaths in Europe. After a thorough pre-analysis of the data, three main statistical models were adapted to the data to pursue the research objectives: the logistic regression model, the classification tree and gradient boosting. A stringency index was also used to compare the situation in the different countries. In commenting on the results, we wanted to focus on the interpretative aspects; in particular, the output of the logistic regression model was interpreted in terms of "odds ratio". The results were interesting for most of the study hypotheses and surprising for some of them, such as the negative influence of contact with children mediated exclusively by technological media.
La presente tesi si propone di studiare gli effetti dell’isolamento conseguente alla prima ondata di Covid-19 sulla salute mentale della popolazione anziana. In particolare, verranno studiati i fenomeni di depressione e solitudine. Tale studio si è basato su un panel di dati raccolti tramite l’indagine SHARE (Survey of Health, Ageing and Retirement in Europe), e, in particolare, su quelli inerenti a SHARE-COVID 19. L'indagine è stata condotta in 28 Paesi diversi, tutti appartenenti all'Unione europea, tranne Svizzera e Israele. L’indagine SHARE nasce per seguire il processo di invecchiamento della popolazione; pertanto, la popolazione target dello studio è composta da persone di età pari o superiore ai 50 anni. Le domande del questionario riguardano diversi ambiti, come la salute fisica e mentale, la situazione economica e le reti sociali degli individui intervistati. In particolare, nell'indagine SHARE-COVID 19 è stata analizzata la situazione prima e dopo la pandemia. Per studiare il fenomeno di interesse, si sono create dunque due variabili risposta, una per la depressione ed una per la solitudine, che rilevano il peggioramento (o l’insorgenza) di queste a causa della pandemia. Si tratta di due variabili di tipo dicotomico, che sono state analizzate separatamente; la distinzione tra le due classi, riguardo appunto ad isolamento e depressione, riguarda chi afferma di sentirsi più depresso e più solo rispetto al periodo pre-pandemico e chi non dichiara di avere avuto questo peggioramento. Partendo dalle domande del questionario, si intendeva capire per quanto possibile quali siano i fattori più significativi che hanno influenzato tali sintomatologie negli anziani. Per esempio, l'uso dei social media ha aiutato gli intervistati a rimanere in contatto con la famiglia? Ha avuto un'influenza positiva sul sentirsi meno soli? Oltre a rispondere a quesiti di questo genere, si è cercato di delineare e scoprire le potenziali differenze tra i Paesi, in base alle rispettive restrizioni adottate. In un primo momento, si è voluto contestualizzare il fenomeno di studio, pertanto si è analizzata la situazione pandemica e la sua evoluzione in termini di infezioni e decessi in Europa. In seguito, dopo aver compiuto un’accurata pre-analisi dei dati, per rispondere agli obiettivi della ricerca si è proceduto con l’adattamento ai dati principalmente di tre modelli statistici: il modello di regressione logistica, l’albero di classificazione e il gradient boosting. Per confrontare la situazione dei diversi Paesi si è inoltre ricorsi ad un indice chiamato “stringency index”. Nel commento dei risultati, ci si è voluti concentrare sugli aspetti interpretativi; in particolare, l’output del modello di regressione logistica è stato interpretato in termini di “odds ratio”. Gli esiti sono stati interessanti per gran parte delle ipotesi di studio e sorprendenti per alcune di queste quali, per esempio l’influsso negativo del contatto coi figli mediato esclusivamente dai supporti tecnologici.
The Covid-19 isolation effect on the mental health of older people
PERONIO, TERESA
2021/2022
Abstract
This thesis aims to study the effects of isolation caused by the first wave of Covid-19 on the mental health of the elderly population. In particular, the phenomena of depression and loneliness will be studied. This study was based on a panel of data collected through the SHARE survey (Survey of Health, Ageing and Retirement in Europe), and mainly on those related to SHARE-COVID 19. The survey was conducted in 28 different countries, all belonging to the European Union, except Switzerland and Israel. The SHARE survey was created to follow the ageing process of the population; therefore, the target population of the study is composed of people aged 50 and over. The questions of the questionnaire cover different fields, such as physical and mental health, economic situation and social networks of the interviewed individuals. In particular, in the SHARE-COVID 19 survey, the situation before and after the pandemic was analysed. To study the phenomenon of interest, two response variables were created, one for depression and one for loneliness, which detect the worsening (or onset) of these due to the pandemic. These are two dichotomous variables, which were analysed separately; for depression and loneliness respectively, the distinction between the two classes concerns those who state that they feel more depressed and lonelier than in the pre-pandemic period and those who do not state that they have experienced this worsening. Starting from the questions in the questionnaire, we intend to understand as far as possible which are the most significant factors that have influenced these symptoms in the elderly. For example, did the use of social media help respondents to stay in touch with their family? Did it have a positive influence on feeling less lonely? In addition to answering these questions, we tried to outline and discover potential differences between the countries according to the respective restrictions adopted. The first step was to contextualise the study phenomenon, and therefore to analyse the pandemic situation and its evolution in terms of infections and deaths in Europe. After a thorough pre-analysis of the data, three main statistical models were adapted to the data to pursue the research objectives: the logistic regression model, the classification tree and gradient boosting. A stringency index was also used to compare the situation in the different countries. In commenting on the results, we wanted to focus on the interpretative aspects; in particular, the output of the logistic regression model was interpreted in terms of "odds ratio". The results were interesting for most of the study hypotheses and surprising for some of them, such as the negative influence of contact with children mediated exclusively by technological media.File | Dimensione | Formato | |
---|---|---|---|
Peronio_Teresa.pdf
accesso aperto
Dimensione
1.7 MB
Formato
Adobe PDF
|
1.7 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/11425