Identifying Factors Associated With Cognitive Impairment in Thyroid Disorders and Prediction of Risk Using Machine Learning Approach: A Comprehensive Study

INTRODUCTION: Thyroid disorders are a significant health concern in India and globally, being the second most common disease. Cognitive impairment associated with thyroid disorders is often neglected by healthcare providers, necessitating improved awareness and screening to identify neurocognitive changes in affected individuals. This study aimed to determine the prevalence of cognitive impairment in thyroid disorders and identify the associated risk factors. This study also predicted the risk of the development of cognitive impairment using a machine learning algorithm and was conducted in a tertiary hospital in South India.
METHODS: A prospective observational study was conducted among 202 patients with thyroid disorders over six months. In order to achieve primary and secondary goals, tools such as the M-ACE questionnaire and XGBoost algorithm were used. Data were collected using a validated form and analyzed using standard statistical methods.
RESULTS: Among the studied population, 29.21% were cognitively impaired, with an average M-ACE Score of 22.56± 2.62. Age, duration of illness, BMI, comorbidities, and tobacco chewing were significant contributors when factors were analyzed for their association with cognitive impairment. A similar number of people with hypothyroidism and hyperthyroidism had a >75% risk of developing cognitive impairment in the near future.
DISCUSSION AND CONCLUSION: This study revealed that thyroid disorders have significant effects on the cognitive status of individuals and were also successful in predicting future risk.