Does there exist an obesity paradox in COVID-19? Insights of the international HOPE-COVID-19-registry

BACKGROUND: Obesity has been described as a protective factor in cardiovascular and other diseases being expressed as ‘obesity paradox’. However, the impact of obesity on clinical outcomes including mortality in COVID-19 has been poorly systematically investigated until now. We aimed to compare clinical outcomes among COVID-19 patients divided into three groups according to the body mass index (BMI). METHODS: We retrospectively collected data up to May 31(st), 2020. 3635 patients were divided into three groups of BMI (<25 kg/m(2); n = 1110, 25−30 kg/m(2); n = 1464, and >30 kg/m(2); n = 1061). Demographic, in-hospital complications, and predictors for mortality, respiratory insufficiency, and sepsis were analyzed. RESULTS: The rate of respiratory insufficiency was more recorded in BMI 25−30 kg/m(2) as compared to BMI < 25 kg/m(2) (22.8% vs. 41.8%; p < 0.001), and in BMI > 30 kg/m(2) than BMI < 25 kg/m(2), respectively (22.8% vs. 35.4%; p < 0.001). Sepsis was more observed in BMI 25−30 kg/m(2) and BMI > 30 kg/m(2) as compared to BMI < 25 kg/m(2), respectively (25.1% vs. 42.5%; p = 0.02) and (25.1% vs. 32.5%; p = 0.006). The mortality rate was higher in BMI 25−30 kg/m(2) and BMI > 30 kg/m(2) as compared to BMI < 25 kg/m(2), respectively (27.2% vs. 39.2%; p = 0.31) (27.2% vs. 33.5%; p = 0.004). In the Cox multivariate analysis for mortality, BMI < 25 kg/m(2) and BMI > 30 kg/m(2) did not impact the mortality rate (HR 1.15, 95% CI: 0.889−1.508; p = 0.27) (HR 1.15, 95% CI: 0.893−1.479; p = 0.27). In multivariate logistic regression analyses for respiratory insufficiency and sepsis, BMI < 25 kg/m(2) is determined as an independent predictor for reduction of respiratory insufficiency (OR 0.73, 95% CI: 0.538−1.004; p = 0.05). CONCLUSIONS: HOPE COVID-19-Registry revealed no evidence of obesity paradox in patients with COVID-19. However, Obesity was associated with a higher rate of respiratory insufficiency and sepsis but was not determined as an independent predictor for a high mortality.