Predicting Depression based on Health-Related Quality of Life (HRQoL)

Download Full-Text PDF Cite this Publication

Text Only Version

Predicting Depression based on Health-Related Quality of Life (HRQoL)

1st Mr.Rahul Pathak

School of Science

MIT World Peace University Pune India

Ms. Akshada Tonape

School of Science

MIT World Peace University Pune, India

Abstract Depression is not only one of the most prevalent of the major psychiatric disorders but also one of the most researched mental illnesses. Previous research has primarily relied on depression detection based on various body and facial expressions, speech prodigy, MRI features and online communities keeping the association limited to external factors. Additionally, very few have contributed in the area of cure for depression. Treating depression effectively means not just keeping it to taking medications and doing therapy. The more changes in lifestyle ensuring a healthy mind and body, the more youll be able to cope with the challenges of depression. Therefore,this survey study focuses on the relationship of food and lifestyle on the depression level of a person. Various health related lifestyle habits with respect to the corresponding HRQOL score can help in finding patterns of habits leading to different depression levels where the depression can be calculated using corresponding questionnaire. This will help in analysing the types of habits to be encouraged contributing to the cure for depression.

Keywords Depression, Mental Health, Machine Learning, Prediction.

  1. INTRODUCTION

    Depression is one of the major risk factors for causing suicide in adolescents, the higher leading cause of death in the specific age group,[4] where more than half suicide victims reported to have a depressive disorder during the time of their death.Depression additionally causes serious social, educational and personal impairments[5],and an increased rate of substance misuse, obesity,smoking and various addictions[8][9].Thus, to recognize the disorder and treatment is important.

    The World Health Organization (WHO) reported the noncommunicable diseases caused by unhealthy lifestyles Various studies have associated between mental health problems in adolescents, eating behaviors, nutrient intake patterns and health related quality of life [2]. Breakfast skipping, high sugar consumptions with soft drinks and sweets were associated with various mental health

    problems.[10],[11] Lifestyle habits such as Sleep Cycle, Breakfast, addictions highly contribute to a persons depression [3]. In the study [1] improvising the diet quality and healthy diets resulted in improvement in mental health of adolescents. A diet with plant foods, fish, regular meals were with better mental health, while nutrient-poor diets and irregular meals resulted into poorer mental health[12],[13].

    Considering the association between (HRQol) and depression and to address the urgent health issue our aim is to predict depression scores based on factors affecting the quality of life in order to analyse the categories of problems to work on in order to have a balanced health.

  2. LITERATURE SURVEY

    Depression is one of the leading cause of disability in people worldwide. Treating and diagnosis of depression have been varied since ages and yet is not up to finding specific solutions for the same. Detection of Depression has been varying and can be a huge step to address the mental illness and offer support to the people suffering from this terrible mental illness.

    Several papers presented a study where depression detection was associated with body movements and expressions.[6] Association of depression detection with facial expressions,vocal atterances and prodigy and speech signal processing. A study additionally focused on sensored data collection to provide diagnosis and detection. Detection being highly based on data collection and availability social media platforms being an integral part of peoples life are used by researchers to identify the causes of depression and detect it . Many studies proposes a text mining approach of using twitter data to detect depression[7]. Detection of depression being a wide area of research focuses on using machine learning methods to predict depression focuses on prediction of suitable treatment.Further in [16],[17] systems for helping provide a solution to depressive patients were developed . Having a few contribution in providing good prediction results for depression our study focuses on achieving good predictions.

    In [2] depression was proved to be associated with nutrition intake,lifestyle habits,behavioural habits[3] where [1] focused on dietry intake patterns being associated with depression.The study [5] focuses on risk factors of depression in adolescence and clinical implications and impairements which results into high suicidal rates in this age group [4]. Various health and diet[8],[11],[10] provides a high contribution of depression rates. Considering the depression

    rates in adolescence and association of lifestyle habits to it our study provides a prediction for depression scores based on HRQoL quotients

  3. PROPOSED SYSTEM

  1. Architecture

  2. Design Considerations.

    • Data Collection and Data Preprocessing:

      Dataset from [2] was used for the survey where data attributes PFSF36, RPSF36, RESF36, VTSF36, MHSF36, SFSF36, BPSF36, GHSF36, EPSSDQ, CPSSDQ, HASSDQ,

      PPSSDQ, PSSSDQ were extracted and used from original dataset. All the null values were replaced with mean .SF36 was used to determine Health related quality of life .The questionnaire consists of 36 questions scaling eight dimensions of lifes quality: Physical Functioning (PF); Role Physical (RP),Bodily Pain (BP), General Health (GH), Vitality (VT), Social Functioning (SF), Role Emotional (RE). Further, Strengths and Difficulties Questionnaires(SDQ) was used for the prediction of mental health issues. SDQ is a screening questionnaire for psychological problems of adolescents. The questionnaire includes 25 questions each of the which is scored on a three pointer type scale with range from 0 to 2. The questions are grouped into a five subscales with 5 questions delivering scores for emotional disorder, hyperactive behavior, peer relationship, pro-social behaviors.

    • Machine Learning

    We used Multi-layer Perceptron regressor for multiple input and multiple output prediction. Features of dataset[2] PFSF36,RPSF36,RESF36,VTSF36,MHSF36,SFSF36,BPSF3

    6,GHSF36 were given to the model to predict EPSSDQ,CPSSDQ,HASSDQ,PPSSDQ,PSSSDQ scores.

    Emotional disorders and peer relationships contribute to internalizing problems and can be further used to predict depression scores. Accuracy was calculated using score function.

    Table of features.

    Prosocial subscale

    PFSF36

    Physical functioning subscale

    RPSF36

    physical difficulties subscale

    RESF36

    emotional difficulties subscale

    VTSF36

    Vitality subscale

    MHSF36

    mental health subscale

    SFSF36

    Social functioning subscale

    BPSF36

    Bodily pain subscale

    GHSF36

    General health subscale

    EPSSDQ

    Emotional problem scale

    CPSSDQ

    Conduct Problem Scale Subscale

    HASSDQ

    Hyperactivity scale subscale

    PPSSDQ

    Peer problems scale

    PSSSDQ

    RESULTS/LIMITATIONS.

    We came up with a novel approach where SDQ scores were predicted for sub parameters based on given HRQoL parameter subscales.Further there are few limitations in our research as the size of the dataset was limited which reduces the accuracy of our prediction. The study focuses on adolescence health stages. The depression score needs to be calculated using two attributes of SDQ scale further.

    CONCLUSION/FUTURE WORK

    Considering Association between health-related quality of life to mental illness, our study provides a prediction for adolescence using neural network. The predication will provide an insight of the quality of lifestyle of adolescence affecting mental health and further inferring on selective and preventive measures to be followed to avoid depression. Our study can further be extended by providing better lifestyle recommendations based on the depression scores.

    ACKNOWLEDGMENTS.

    We would like to thank our refree Mrs. Supriya Aras. For guiding us throughout the research process. This paper wouldnt have been a success without her involvement for the same.

    REFERENCES

    1. Jacka FN, Kremer PJ, Berk M, Silva-Sanigorski AM, Moodie M, Leslie ER, et al. A prospective study of diet quality and mental health in adolescents. Plos One. 2011; 6: E24805. https://doi.org/10.1371/ journal.pone.0024805 PMID: 21957462

    2. Farhangi M.A., Dehghan P., Jahangiry L. Mental health problems in relation to eating behavior patterns, nutrient intakes and health related quality of life among Iranian female adolescents. PLoS One. 2018;13(4) Article number e0195669

    3. Frederick, T., Frerichs, R. R., & Clark, V. A. (1988). Personal health habits and symptoms of depression at the community level. Preventive Medicine, 17(2), 173182.doi:10.1016/0091- 7435(88)90061-8

    4. Windfuhr K, While D, Hunt I, et al. National confidential inquiry into suicide and homicide by people with mental illness. Suicide in juveniles and adolescents in the United Kingdom. J Child Psychol Psychiatry. 2008;49:115565.

    5. Lewinsohn PM, Rohde P, Seeley JR. Major depressive disorder in older adolescents: prevalence, risk factors and clinical implications. ClinPsychol Rev. 1988;18:76594

    6. Rosa, R. L., Rodriguez, D. Z., Schwartz, G. M., de Campos Ribeiro, I., & Bressan, G. (2016). Monitoring system for potential users with depression using sentiment analysis. 2016 IEEE International Conference on Consumer Electronics (ICCE). doi:10.1109/icce.2016.7430656

    7. Sau, A., & Bhakta, I. (2017). Predicting anxiety and depression in elderly patients using machine learning technology . Healthcare Technology Letters, 4(6), 238243.

      doi:10.1049/htl.2016.0096

    8. Keenan-Miller D, Hammen CL, Brennan PA. Health outcomes related to early adolescent depression. J Adolesc Health. 2007;41:25662.

    9. Hasler G, Pine DS, Kleinbaum DG, et al. Depressive symptoms during childhood and adult obesity: the Zurich cohort study. Mol Psychiatry. 2005;10:84250

    10. Overby N, Hoigaard RNA. Diet and behavioral problems at school in Norwegian adolescents. Food Nutr Res. 2012, 56:1723117237. 13.

    11. Liena L. Is breakfast consumption related to mental distress and academic performance in adolescents? Public Health Nutr 2007; 10(4): 422428. https://doi.org/10.1017/S1368980007258550 PMID: 17362539

    12. Oellingrath IM, Svendsen MV, Hestetun I. Eating patterns and mental health problems in early adolescence–a cross-sectional study of 12-13-year-old Norwegian schoolchildren. Public Health Nutr. 2014;17(11):25542562. doi:10.1017/S1368980013002747

    13. Keenan-Miller D.HammenCL,Brenan PA(2007) Health outcomes related to early adolesents depression. J ADolsecHelth 41:225-231

    14. Ghanizadeh A, Izadpanah A. Scale Validation of the Strengths and Difficulties Questionnaire in Iranian Children. Iran J Psychiatry 2007; 2: 65-71.

    15. Xuan J, Kirchdoerfer LJ, Boyer JG, Norwood GJ (1999) Effects of comorbidity on health-related quality-of-life scores: an analysis of clinical trial data. Clin Ther 21:383403.

    16. Yang, S., Zhou, P., Duan, K., Hossain, M. S., & Alhamid, M. F. (2017). emHealth: Towards Emotion Health Through Depression Prediction and Intelligent Health Recommender System. Mobile Networks and Applications, 23(2), 216226. doi:10.1007/s11036- 017-0929-3

    17. Chen, M., Zhang, Y., Qiu, M., Guizani, N., & Hao, Y. (2018). SPHA: Smart Personal Health Advisor Based on Deep Analytics. IEEE Communications Magazine, 56(3), 1

Leave a Reply

Your email address will not be published. Required fields are marked *