Skip to main content

Body Mass Index as a Risk Factor for Clinical Outcomes in Patients Hospitalized with COVID-19 in New York

Our take —

This large retrospective cohort study from the early phase of the pandemic (March 1 to May 12, 2020) found that among 10,861 inpatients in New York who tested positive for SARS-CoV-2 infection, underweight and obese patients were at greater risk of death, whereas overweight and obese patients were at greater risk of invasive mechanical ventilation. However, among patients who received invasive mechanical ventilation, there was no association between BMI and death. These findings should be interpreted with caution given probable selection bias and residual confounding.

Study design

Retrospective Cohort

Study population and setting

This study retrospectively investigated the relationship between body mass index (BMI) categories and COVID-19 morbidity and mortality outcomes using electronic medical records of adult inpatients admitted to one of 12 Northwell Health system hospitals in New York state between March 1 and April 27, 2020, who tested positive for SARS-CoV-2 infection with RT-PCR (n=10,861). Multivariable regression models were used to adjust for patient characteristics including age, sex, race/ethnicity, comorbidities, smoking status, hospital type, and BMI groups. Primary outcomes included invasive mechanical ventilation and inpatient death on or before May 12, 2020. The study population was 59.6% male with a median age of 65 years (interquartile range [IQR] = 54-77 years). The racial and ethnic composition was 33.8% white, 20.9% Black, 22.8% Hispanic, and 8.6% Asian.

Summary of Main Findings

The patient BMI distribution is as follows: 2.2% underweight (<18.5 kg/m2), 23.1% normal (18.5-24.9 kg/m2), 37.0% overweight (25-29.9 kg/m2), 21.6% obesity class I (30-34.9 kg/m2), 9.1% obesity class II (35-39.9 kg/m2), and 7.0% obesity class III (≥40 kg/m2). In adjusted analyses using normal BMI as the reference group, overweight and each obesity class was associated with increased odds of invasive mechanical ventilation: overweight (adjusted odds ratio [AOR]=1.27; 95% confidence interval [CI]: 1.11-1.46), obesity class I (AOR=1.48; 95% CI: 1.27-1.72), obesity class II (AOR=1.89; 95% CI: 1.56-2.28), and obesity class III (AOR=2.31; 95% CI: 1.88-2.85). Underweight (AOR=1.44; 95% CI: 1.08-1.92), obesity class II (AOR=1.25; 95% CI: 1.03-1.52), and obesity class III (AOR=1.61; 95% CI: 1.30-2.00) were associated with increased odds of inpatient death. Among patients placed on invasive mechanical ventilation, no BMI category was associated with inpatient death.

Study Strengths

This large multicenter cohort study included a racial and ethnically diverse study population representative of groups most heavily impacted by COVID-19 in urban settings. Patient BMI categorizations were clearly defined, standardized, and assessed. This study adjusted for covariates which may account for some confounding.


This study looked at an exclusive inpatient population localized to highly populated urban and suburban areas in New York state, limiting generalizability to outpatient settings, persons with milder disease, or other geographical settings. Since this study took place early in the pandemic when COVID-19 testing resources were limited and clinical diagnosis was less standardized, selection bias could have existed and selected for patients with more severe or classic disease presentations. Due to the use of data from a single administrative system, this study could not account for outcomes of discharged patients or patients who subsequently sought care at other institutions which may have resulted in an underestimate of invasive mechanical ventilation and death rates. This study could not provide insight into the clinical decision-making process, and it could not rule out the possibility of clinician bias toward early intervention, such as invasive mechanical ventilation, in overweight and obese patients may explain some of the observed association. This study was not randomized, and residual confounding is likely to exist. Additionally, it is possible that over-adjustment for some confounders, such as disease comorbidities that may mediate causation between BMI and COVID-19 disease outcomes, could have made associations appear weaker.

Value added

This study is among the first to elucidate the relationship between continuity of BMI categories and severe COVID-19 outcomes in a large cohort.

This review was posted on: 14 January 2021