Correlation between obesity, age and mortality for COVID-19 patients with acute respiratory distress syndrome supported by extracorporeal membrane oxygenation

To the editor,

There were wide ranges of mortality rates (15%∼65%) for COVID-19 patients with acute respiratory distress syndrome (ARDS) treated by extracorporeal membrane oxygenation (ECMO).¹ Some report claimed obesity was a protective factor for the COVID-19 patients with ECMO support,² whereas obesity was also thought as a risk factor for COVID-19 infection.^{3 , 4} This systematic review was performed with an attempt to clarify the controversy.

We performed a systemic literature search in PubMed, EMBASE, Web of Science and MEDLINE. Key words used for searching included “COVID”, “coronavirus”, “ARDS”, and “ECMO”. The searched articles were included in this study if the following criteria were met: (1) adult patients; (2) COVID-19 infection; (3) ARDS status; (4) ECMO use. Review articles were excluded. Exclusion criteria were listed as mentioned below: (1) pediatric patients; (2) data overlap; (3) duplicated publication; (4) non-English articles.

By our searching strategy, 1545 studies were retrieved from the databases. After screening, 72 articles were included in this study (supplement file). Among the 72 articles, there were 8540 COVID-19 patients treated with ECMO and the publication year was from 2020 to 2022. The age of the patients was reported as mean age in 23 articles (4748 patients; pooled mean age, 51.2 years; range of mean age, 31.1–64.3 years), was reported as median age in 47 articles (3758 patients, range of median age 39.0–64.5 years old), was reported as age range by 1 article and not reported in 1 article. The body mass index (BMI) was reported as mean BMI in 13 articles (286 patients, pooled mean BMI 32.5 kg/m2, range of mean BMI 24.2–35.1 kg/m2), was reported as median BMI in 34 articles (3018 patients, range of median BMI 23.7–34.7 kg/m2), was reported as BMI>30 in 3 articles (2034 patients, pooled percentage for BMI>30: 61.7%) and not reported in 22 articles.

The pooled in-hospital mortality rate was 48.5% (2918/6020) for the searched studies in which the ECMO patients completed their hospital treatment and were discharged from hospital either dead or alive. Regarding the correlation with in-hospital mortality (supplement file), the correlation coefficient was 0.6481 (p = 0.0002) for median age and −0.4324 (p = 0.064) for median BMI. There was a significantly negative correlation between median age and median BMI (correlation coefficient −0.6216, p = 0.0001).

“Obesity paradox”² was ever described for the patients with pneumonia before and during COVID-19 pandemic. However, this viewpoint seems a little bit confusing because it has been known that obesity is associated with developing severe disease³ and death⁴ from COVID-19 infection. According to our study (Fig. 1 ), there was significantly positive correlation between median age and in-hospital mortality, and significantly negative correlation between median age and median BMI, which could explain why there was a trend of negative correlation between median BMI and in-hospital mortality.

Fig. 1.

Scatter plots to present (A) the correlation (p = 0.0002) between median age and in-hospital mortality; (B) the correlation (p = 0.064) between median BMI and in-hospital mortality; (C) the correlation (p = 0.0001) between median age and median BMI (one circle indicates one study; the circle size is weighted by the patient number in each article included in our study).

In conclusion, for the COVID-19 patients treated with ECMO, age was an independent factor for in-hospital mortality, while obesity was not a protective factor for in-hospital mortality as the patients with younger age was associated with higher BMI.

Declaration of competing interest

No financial and non-financial conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article.