“…Therefore, relevant variables were identified based on professional knowledge ( 15 , 16 ), and their associations with the self-reported duration of severe symptoms of COVID-19 were examined as the main outcome. Employing a historical confounding approach, a multivariate regression model incorporated age (years) ( 17 ), BMI group ( 18 ), education level ( 19 ), marital status ( 20 ), smoking status ( 21 ), alcohol consumption ( 22 ), bariatric surgery type (SG, RYGB, and others) ( 23 ), hypertension ( 24 ), and diabetes ( 25 ). Due to the skewed distribution of BMI data, we grouped the values into intervals with uniform spacing of 5, namely (<20, 20–25, 25–30, 30–35, 35–40, 45–50, 50–55, and 55–60 kg/m 2 ).…”