Dissertation

Modelling of Longitudinal Digital Health Data to Understand Underlying Phenotypes

Das, Rajenki

2023

Overview

Health is an integral part of human well being, and a holistic perspective on the same involves physical, mental and social factors (World Health Organization et al., 1948). However, “health” itself is a complex concept where many aspects play a role in building a healthy person and meanings of the same can vary across individuals, as per the capability of tackling an illness (Leonardi, 2018). Most of us are aware of physical illness and associated concerns but mental health often gets neglected in the larger discourse on health and wellbeing, and it remains important to remember the adage “no health without mental health” as used used by e.g. Prince et al. (2007) in the context of overall health. Even though recently there has been an increase in awareness towards mental health, the stigma around it continues to exist (Bharadwaj et al., 2017; Gold et al., 2016). Additionally this social stigma around mental health as well as a self-perceived notion of underestimating an issue (Andrade et al., 2014) often leads to lack of treatments. Proper treatments or diagnoses are still unavailable or inaccessible for many individuals (Moreno et al., 2020; Camm-Crosbie et al., 2019; MacDonald et al., 2018). In 2019, the World Health Organisation (WHO) estimated 970 million people in the world to be living with a mental disorder (Organization, 2022). Keeping all of this in mind, it needs to be reiterated that mental ailments are widespread and can affect anyone, just like a physical illness. There are many aspects of this topic that need to be dealt with carefully, but in this thesis, the emphasis has been on finding mental health traits using digital health records and how it can facilitate the process of developing treatments for the same.Mental health can be affected by numerous factors, with examples being: physical health problems, socio-economic conditions, nutrition, genetics, and the environment around us (Kola et al., 2021; Adan et al., 2019; Bhugra et al., 2013; Tew et al., 2012; Rutter, 2005; Morris, 2003; Tsuang, 2000). Even though identifying causes remains an extremely challenging problem, symptoms associated with a decline in mental health can aid diagnoses. It can be found that withdrawal in life, many times shown by lack of motivation, is quite prominent amongst those severely affected by mental affliction and can serve as a vital warning. But, a more comprehensive understanding of mental health requires an interdisciplinary approach (Fried, 2021) which can benefit from insights from various fields including neurology, psychology, sociology, biology etc. and, importantly in the current context, mathematical sciences. The advent of COVID-19 presented a global health crisis which affected lives across the world quite disproportionately (Gibson et al., 2021) and resulted in further widening of inequalities. A rise in mental health problems, as an associated result of worsening physical health or in this case, a physical health calamity, has become a major concern (Vigo et al., 2020) and may have long-term implications (Bourmistrova et al., 2022; Kola et al., 2021). The motivation behind this thesis was to help quantify mental health, model it and see underlying behaviour using mathematical, statistical and computational tools as these would help in comparing and providing better tools for understanding the differences and commonalities in behaviour patterns.While there are pros and cons to the advancement of technology and electronic health (eHealth) (Vitacca et al., 2009), in this context, it has provided us with digital health tools which have facilitated the collection of information on health. Mobile health (mHealth) is potentially revolutionary and opening up doors to opportunities for exploring research in healthcare (Fiordelli et al., 2013). Especially in the sphere of mental health, mhealth helps in overcoming many barriers related to accessibility (Price et al., 2014). Such apps or platforms can be beneficial to those who are unable or are reluctant to be available for an in person appointment, as well as those who want to keep their information completely private or anonymous. This is particularly true in the context of mental health, where many prefer not to be identified while reporting their issues as a result of the stigma attached, although it is difficult to tell if a mHealth based solution or therapy is indeed better than an in-person one (Olff, 2015) but nevertheless, these apps can be useful as they maintain the confidentiality of the patient.

• Publisher: ProQuest Dissertations & Theses
• Subject: Chronic pain / Markov analysis / Mathematics / Medicine / Normal distribution / Operations research / Probability / Self report / Sleep
• ISBN: 9798381847680