Skip to main content
NCBI home page
Elsevier - PMC COVID-19 Collection logoLink to Elsevier - PMC COVID-19 Collection
. 2022 Aug 6;51:377–384. doi: 10.1016/j.clnesp.2022.07.016

Body mass index and clinical outcome of severe COVID-19 patients with acute hypoxic respiratory failure: Unravelling the “obesity paradox” phenomenon

Michael Jennings a,b, Maria Burova a, Laura G Hamilton a, Elsie Hunter a, Clare Morden a, Darshni Pandya a, Ryan Beecham a,b, Helen Moyses b, Kordo Saeed c,d, Paul R Afolabi b,d, Philip C Calder b,d, Ahilanandan Dushianthan a,b,d,∗; the REACT COVID-19 Investigators
PMCID: PMC9356629  PMID: 36184231

Abstract

Background and aims

Although obesity have been generally shown to be an independent risk factor for poor outcomes in COVID-19 infection, some studies demonstrate a paradoxical protective effect (“obesity paradox”). This study examines the influence of obesity categories on clinical outcomes of severe COVID-19 patients admitted to an intensive care unit with acute hypoxic respiratory failure requiring either non-invasive or invasive mechanical ventilation.

Methods

This is a single centre, retrospective study of consecutive COVID-19 patients admitted to the intensive care unit between 03/2020 to 03/2021. Patients were grouped according to the NICE Body Mass Index (BMI) category. Admission variables including age, sex, comorbidities, and ICU severity indices (APACHE-II, SOFA and PaO2/FiO2) were collected. Data were compared between BMI groups for outcomes such as need for invasive mechanical ventilation (IMV), renal replacement therapy (RRT) and 28-day and overall hospital mortality.

Results

340 patients were identified and of those 333 patients had their BMI documented. Just over half of patients (53%) had obesity. Those with extreme obesity (obesity groups II and III) were younger with fewer comorbidities, but were more hypoxaemic at presentation, than the healthy BMI group. Although non-significant, obesity groups II and III paradoxically showed a lower in-hospital mortality than the healthy weight group. However, adjusted (age, sex, APACHE-II and CCI) competing risk regression analysis showed three-times higher mortality in obese category I (sub-distribution hazard ratio = 3.32 (95% CI 1.30–8.46), p = 0.01) and a trend to higher mortality across all obesity groups compared to the healthy weight group.

Conclusions

In this cohort, those with obesity were at higher risk of mortality after adjustment for confounders. We did not identify an “obesity paradox” in this cohort. The obesity paradox may be explained by confounding factors such as younger age, fewer comorbidities, and less severe organ failures. The impact of obesity on indicators of morbidity including likelihood of requirement for organ support measures was not conclusively demonstrated and requires further scrutiny.

Keywords: COVID-19, Intensive care, Body mass index, Obesity, Acute hypoxic respiratory failure, Invasive mechanical ventilation

1. Introduction

Over the last two years, the COVID-19 pandemic, caused by a novel strain of beta coronavirus (SARS-CoV-2), has secured its place in history, infecting over 300 million people and placing unprecedented stress on the global health infrastructure [1]. The emergence of this pandemic has focused attention on individual risk stratification, in part, because constraints on healthcare resources such as immunisations have promoted a strategy of protecting the most vulnerable individuals first. The risk factors for increased severity of COVID-19 infection are well-documented and include increased age, ethnicity, presence of cardiovascular disease, chronic lung disease and diabetes mellitus [[2], [3], [4]].

The role of obesity in patient outcomes is of interest due to its increasing prevalence across western and now developing world. Based on global data from 2016, the it is estimated that by 2025, global obesity prevalence will reach 18% in men and surpass 21% in women [5]. There is also a large body of evidence that supports obesity as a risk factor for poorer outcome in a range of respiratory viral infections, including H1N1 influenza and the Middle East Respiratory Syndrome–Related Coronavirus (MERS-CoV) [6,7]. The impact of obesity on the disease course of COVID-19 is less clear, with outcome heterogeneity across a range of observational studies. A large multi-centre retrospective cohort study of critically ill COVID-19 patients demonstrated a linear association between BMI and the need for invasive mechanical ventilation, independent of other metabolic risk factors, and a non-linear association between increasing BMI and 28-day all-cause mortality [8]. This relationship between BMI and mortality was also investigated by several other observational studies [[9], [10], [11], [12], [13]], but the relationship was not universally observed. Some studies found increased BMI was associated with an increased risk of requiring intubation and ventilation, but with no clear relationship with mortality [[14], [15], [16], [17]]. In contrast, the Intensive Care National Audit and Research Centre (ICNARC) highlighted a paradoxical relationship within the national dataset whereby increasing BMI was associated with lower 28-day in-hospital mortality [18]. This finding is supported by the Practice of Ventilation in COVID-19 (PRoVENTCOVID) study, which enrolled patients from 22 intensive care units and found higher rates of in-hospital mortality, ICU mortality and 90-day all-cause mortality in overweight and healthy individuals than obese patients. However, this effect diminished when the analysis was adjusted for age, co-morbidities, and a range of other physiological parameters [19].

The “Obesity Paradox”, where higher BMI can appear protective in individual datasets, has been previously observed in end-stage renal disease [20], chronic obstructive pulmonary disease (COPD) [21], and pulmonary embolism [22]. If there is an “obesity paradox” in the context of COVID-19 [23], its root cause is not universally agreed upon. Consequently, we explored the relationship between body mass index (BMI) and outcomes in a retrospective cohort of critically ill COVID-19 patients admitted to our intensive care unit with acute hypoxic respiratory failure.

2. Methods

All adults (≥18 years old) in whom SARS-CoV-2 positivity was confirmed by RT-PCR (Reverse-Transcriptase Polymerase Chain Reaction) admitted to the General Intensive Care Unit (GICU) at University Hospital Southampton NHS Foundation Trust between 19th March 2020 and 12th March 2021, a period that covered two distinct waves of the pandemic in the UK, were included in this study. Patients were sub-categorized into five BMI classes according to National Institute of clinical Excellence (NICE) BMI classification: healthy weight (BMI 18.5–24.9 kg/m2), overweight (25–29.9 kg/m2), Obesity I (30–34.9 kg/m2), Obesity II (35–39.9 kg/m2) and Obesity III (>40 kg/m2) [24]. This study has ethical approval as part of “a longitudinal cohort study to facilitate better understanding and management of SARS-CoV-2 infection from hospital admission to discharge across all levels of care (REACT-COVID19)” (North West Research Ethics Committee (REC 17/NW/0632), and Specific Review Board (REC 20/HRA/2986)) [25]. Taking of consent was waived due to the nature of the study. Collected data were anonymised and handled according to the local institutional and national policies. The study used STROBE guidelines for reporting observational studies [26].

Demographic data including age, sex, clinical frailty score (CFS), and comorbidities (type-I diabetes, type-II diabetes, asthma, COPD, hypertension, ischemic heart disease, congestive cardiac failure, arrhythmia, chronic kidney disease, immunosuppression, and pre-existing steroid use) were all collected. Comorbidities are quantitatively presented as Charlson's Comorbidity Index (CCI) [27]. Admission severity indices including Acute Physiology and Chronic Health Evaluation (APACHE-II), Sequential Organ Failure Assessment (SOFA) scores, oxygenation indices with laboratory variables were documented during the first 24 h of admission. The ICU organ support measures including the use of non-invasive ventilation (NIV), invasive mechanical ventilation and renal replacement therapy were collected. The outcomes assessed were 28-day and overall hospital mortality and were available for all patients. Clinical management of patients was in accordance with local guidelines. Most patients admitted to the ICU required ventilatory support either in the form of NIV or mechanical ventilation. Our local guidance suggests the initiation of NIV in all patients with severe acute hypoxic respiratory failure needing inspired oxygen of FiO2 >60%. The type of NIV initiated was at the discretion of the treating physician. Deteriorating patients were subsequently intubated and mechanically ventilated.

Continuous variables are reported as median and interquartile range (IQR) and the group comparisons are made using Kruskal–Wallis test. Categorical variables are reported as number and percentage and analysed using Fisher's exact test. Significance was defined at p < 0.05. Pearson's correlation coefficient was used to assess the relationship of demographic variables and BMI. Logistic regression analysis was used to assess the risk of invasive mechanical ventilation and renal replacement therapy according to the BMI categories. Unadjusted and adjusted competing risk regression analyses (adjusted for age, sex, CCI and APACHE II score) were carried out to assess the effect of BMI categories on 28-day and overall hospital mortality and presented as the sub-distribution hazard ratio (SHR). Statistical analysis was performed using SPSS version 27, Stata version 17.0, R version 1.2.5042 and MedCalc version 20.008.

3. Results

During the period between March 2020 and March 2021, there were 340 patients admitted to the General Intensive Care Unit due to COVID-19 related pneumonia. SARS-CoV-2 infection was confirmed by RT-PCR in all patients. Of those admitted, 66% were male with a median age of 59 years (IQR 49, 69) and 77% of white ethnic origin. The median days between symptom onset to hospitalization was seven (IQR 4, 10). Patients were categorized according to the National Institute of Clinical Excellence, UK (NICE) obesity classification groups as healthy, overweight, obesity I, obesity II and obesity III, accounting for 17.6%, 27.4%, 25.6%, 15.6% and 11.8% of patients respectively (Fig. 1 ). The median BMI was 31 kg/m2 (IQR 26,35). The admission demographics for individual BMI categories are presented in Table 1 .

Fig. 1.

Fig. 1

Open in a new tab

Prisma flow chart displaying the proportion of patients according to BMI category.

Table 1.

Characteristics of subjects with severe COVID-19 according to thier. Each parameter is expressed as median (interquartile range) or number (percentage). Comparisons are made using Kruskal–Wallis test (continuous variables) or Fisher's exact test (categorical variables). ∗p < 0.05. ACEi: Angiotensin converting enzyme inhibitors; APACHE-II: Acute Physiology and Chronic Health Evaluation II score; ARB: Angiotensin Receptor Blockers; BMI: body mass index; COPD: Chronic Obstructive Pulmonary Disease; LDH: Lactate dehydrogenase; SOFA: sequential organ failure assessment score.

Demographics Healthy BMI 18.5–24.9 (N = 60) Overweight BMI 25–29.9 (N = 93) Obesity I
BMI 30–34.9 (N = 87)		 Obesity II
BMI 35–40 (N = 53)		 Obesity III
BMI ≥40 (N = 40)		 Group Significant
Age, years 64 (51,75) 62 (52,72) 60 (49,67) 57 (48,61) 49 (43,59) ∗Yes
Male (%) 65% 82% 61% 68% 45% ∗Yes
Symptomatic days prior to hospitalisation 7 (4,10) 8 (4) 7 (4,9) 7 (5,10) 7 (4,9) No
BMI (kg/m2) 23 (22,24) 27 (26,28) 32 (31,34) 37 (36,39) 44 (42,51) ∗Yes
Clinical frailty score 2 (1,3) 2 (2,3) 2 (2,3) 2 (2,4) 2 (2,3) No
Charlson's comorbidity index 3 (1,4) 3 (1,4) 2 (1,3.5) 2 (1,3) 1 (0,2) ∗Yes
Race/ethnic group
White 40 (66.7%) 67 (72.0%) 69 (79.3%) 46 (86.8%) 31 (77.5%)
Black 4 (6.7%) 4 (4.3%) 4 (4.6%) 3 (5.7%) 2 (5.0%) No
Asian/Indian 12 (20%) 19 (20.4%) 12 (13.8%) 3 (5.7%) 5 (12.5%)
Mixed/Other/unknown 4 (6.7%) 3 (3.3%) 2 (2.2%) 1 (1.9%) 2 (5%)
Comorbidities, n (%)
Asthma 3 (5%) 12 (12.9%) 8 (9.2%) 8 (15.1%) 8 (20.0%) No
COPD 1 (1.7) 7 (7.5%) 7 (8.0%) 2 (3.8%) 2 (5.0%) No
Chronic kidney disease 5 (8.3%) 9 (9.7%) 3 (3.4%) 2 (3.8%) 4 (10.0%) No
Congestive cardiac failure 0 (0%) 11 (11.8%) 3 (3.4%) 0 (0%) 2 (5.0%) ∗Yes
Diabetes mellitus 11 (18.3%) 26 (28%) 28 (32.2%) 17 (32.1%) 16 (40.0%) No
Hypertension 23 (38.3%) 43 (46.2%) 35 (40.2%) 22 (41.5%) 15 (37.5%) No
Ischemic heart disease 9 (15%) 19 (20.4%) 7 (8.0%) 3 (5.7%) 4 (10.0%) ∗Yes
Immunosuppression 7 (11.7%) 11 (11.8%) 2 (2.3%) 5 (9.4%) 2 (5.0%) No
Presence of another comorbidity 32 (53.3%) 59 (63.4%) 55 (63.2%) 33 (62.3%) 25 (62.5%) No
Use of ACEi or ARB 13 (21.7%) 33 (35.5%) 28 (32.2%) 9 (17.0%) 16 (40.0%) ∗Yes
ICU Severity indices
APACHE II Score 15 (11,23) 14 (10,20) 14 (9,14) 11 (8,15) 11 (9,16) ∗Yes
SOFA Score 4 (3,6) 4 (3,6) 4 (3,4) 4 (3,5) 4 (3,5) No
PaO2/FiO2 ratio (kPa) 15.1 (12.5,21.1) 14.6 (12.7,19.5) 14.0 (11.3,17.9) 14.0 (11.5,16.3) 11.1 (8.8,14.7) ∗Yes
ICU interventions
Mechanical ventilation, n (%) 38 (63.3%) 50 (53.8%) 43 (49.4%) 29 (54.7%) 19 (47.5%) No
Non-invasive ventilation, n (%) 17 (28.3%) 39 (41.9%) 38 (43.7%) 22 (41.5%) 21 (52.5%) No
Renal replacement therapy, n (%) 12 (20%) 14 (15%) 12 (13.8%) 11 (20.8%) 3 (7.5%) No
Admission laboratory profile
Bilirubin (mmol/l) 10 (8,16) 11 (8,13) 10 (8,14) 11 (8,13) 9 (7,12) No
Creatinine (mmol/l) 73 (58,107) 70 (58,96) 69 (54,92) 73 (64, 96) 63 (51,105) No
Creatinine kinase (U/l) 117 (54,311) 134 (73,372) 122 (60,332) 158 (82,474) 217 (94,519) No
C-Reactive Protein (mg/l) 124 (73,209) 104 (51,185) 125 (80,189) 113 (68,169) 144 (76,182) No
D-Dimer (mg/l) 765 (366,2010) 694 (731,1302) 452 (267, 909) 485 (233,773) 400 (278,1069) ∗Yes
Ferritin (mg/l) 776 (379,1709) 883 (538,1286) 652 (390, 1177) 851 (527, 1402) 318 (160,645) ∗Yes
HbA1c (mmol/mol) 48 (43,54) 44 (40,57) 47 (41,54) 49 (43,60) 43 (40,61) No
LDH (U/l) 955 (645,1199) 946 (731, 1302) 855 (695,1209) 909 (793,1235) 961 (766,1270) No
Lymphocytes 109/l 0.7 (0.6,0.9) 0.7 (0.4,1.1) 0.8 (0.6, 1.2) 0.7 (0.6,0.9) 0.8 (0.6,1.1) No
Neutrophil/lymphocyte ratio 10.5 (5.3,15.4) 10.1 (6.3,18.0) 8.7 (5.8, 14.4) 10 (6.7,15.5) 7.6 (5.0, 12.0) No
Procalcitonin (ng/ml) 0.2 (0.1,1.0) 0.3 (0.1,0.9) 0.3 (0.1, 0.7) 0.2 (0.1,0.5) 0.2 (0.1,0.4) No
HS Troponin (ng/l) 16 (11,94) 20 (9,53) 10 (7,34) 11 (7,20) 20 (5.5,59) ∗Yes
White cell counts 109/l 9.4 (6,12) 8.5 (6.2,11.8) 8.9 (6.6,12.8) 8.3 (5.8,11.0) 7.3 (5.4,8.7) No
Open in a new tab

Those in obesity groups II and III were younger than the healthy BMI group. The proportional distribution of obesity categories between sexes is pictorially presented as Fig. 2 . There were no differences in the timing of symptomatic presentation to hospital between groups. The weighted composite score for comorbidities (Charlson's Comorbidity Index) was lower in obesity groups II and III in comparison to the overweight or healthy categories with individual differences in the presence of congestive cardiac failure and ischemic heart disease. Although there were no differences in the SOFA scores at admission between groups, the APACHE-II scores were significantly lower in the obesity II and III groups compared to the healthy BMI group. Moreover, the admission oxygen index defined by the ratio of arterial oxygen (PaO2) to inspired fractional oxygen (FiO2) was lower in patients in obesity group III, suggesting more severe hypoxemia at presentation in this group. Admission laboratory variables were similar except for ferritin, D-Dimer, and cardiac troponin levels between groups (Table 1).

Fig. 2.

Fig. 2

Open in a new tab

Population pyramid displaying number of male vs female patients according to their BMI.

We assessed the relationship between BMI and admission variables (age, APACHE-II and PaO2/FiO2 ratio) which demonstrated group differences. There was a negative correlation between BMI and age (Pearson r = −0.24, p < 0.001), BMI and APACHE-II score (Pearson r = −0.23, p < 0.001), and BMI and PaO2/FiO2 ratio (Pearson r = −0.25, p < 0.001). This suggests that patients with a higher BMI are more likely to be young, with single organ failure/less chronic comorbidity (defined by lower APACHE II score) but with increased severity of hypoxemia at presentation (Fig. 3 ). Overall, 54.1% of patients required mechanical ventilation and 15.6% developed acute kidney injury and needed renal replacement therapy (RRT). We performed logistic regression analysis for these outcomes of requirement for mechanical ventilation or RRT using the BMI groups adjusting for covariates including age, sex, CCI and APACHE-II scores. There were no significant differences in the need for mechanical ventilation or RRT between BMI category groups in this analysis (Table 2 ).

Fig. 3.

Fig. 3

Open in a new tab

Scatter plots for correlation between BMI with Age (A), APACHE-II score (B) and oxygenation (PaO2/FiO2) on admission (C).

Table 2.

Logistic regression analysis for the outcomes need for mechanical ventilation and need for renal replacement therapy according to BMI groups adjusted for age, sex, Charlson's Comorbidity Index (CCI) and APACHE-II. Healthy BMI group is used as reference.

BMI category Outcome n, (%) Odds ratio (95% CI) p-value
Mechanical ventilation
Healthy 38 (63.3%) 1.000 N/A
Overweight 50 (53.8%) 1.037 (0.472–2.279) 0.928
Obesity I 43 (49.3%) 0.828 (0.370–1.851) 0.645
Obesity II 29 (54.7%) 1.594 (0.047–3.926) 0.310
Obesity III 19 (47.5%) 0.845 (0.321–2.227) 0.733
Renal replacement therapy
Healthy 12 (20.0%) 1.0 N/A
Overweight 14 (15.1%) 0.776 (0.305–1.977) 0.596
Obesity I 12 (13.8%) 0.985 (0.382–2.542) 0.975
Obesity II 11 (20.8%) 2.139 (0.763–6.002) 0.148
Obesity III 3 (7.5%) 0.508 (0.119–2.175) 0.362
Open in a new tab

Of those patients with BMI measurements (N = 333) and subdivided according to the BMI category, there were no significant differences in the duration of ICU or hospital length of stay between the groups. The 28-day and overall hospital mortality were 18.3% and 23.7% respectively. The 28-day hospital mortality for the groups healthy, overweight, obesity I, obesity II, obesity III was 21.7%, 28.0%, 21.8%, 17.0% and 12.5% respectively. A similar trend was also seen for the overall hospital mortality which was 21.7%, 30.1%, 26.4%, 17.0% and 15.0% respectively (Table 3 ). Competing-risks regression models were used to assess the impact of obesity categories on mortality outcomes, 28-day in-hospital mortality and overall hospital mortality with discharge from hospital specified as a competing risk. Both unadjusted and adjusted (age, sex, APACHE-II, and CCI) models were utilised. Although it appeared that obesity groups II and III had lower 28-day and overall hospital mortality, these were not statistically significant (Table 4 and Fig. 4 ). In each analysis, the sub-distribution hazard ratio (SHR) determines the risk of death for each BMI category comparing to the healthy-weight category after adjusting for the confound covariates. All obesity groups had greater 28 day and overall hospital mortality, but it was only statistically significant for BMI group I [SHR 3.31 (p = 0.01) and SHR 2.82 (p = 0.009)] respectively than healthy BMI patients.

Table 3.

Intensive Care Unit and hospital outcomes according to BMI group. Continues variables analysed by Kruskall–Wallis test and categorical variables by Fisher's exact test using Chi-square statistic.

Healthy
N = 60		 Overweight
N = 93		 Obesity I
N = 87		 Obesity II
N = 53		 Obesity III
N = 40		 p- value
Duration of ICU length of stay (days) 11 (3,34) 11 (6,23) 9 (4,19) 13 (7,20) 10 (5,22) 0.59
Duration of hospital length of stay (days) 22 (14,49) 19 (12,38) 22 (11,35) 21 (12,28) 18 (10,37) 0.30
28-day survival n, (%) 13 (21.7%) 26 (28.0%) 19 (21.8%) 9 (17.0%) 5 (12.5%) 0.30
Overall hospital survival n, (%) 13 (21.7%) 28 (30.1%) 23 (26.4%) 9 (17.0%) 6 (15.0%) 0.23
Open in a new tab

Table 4.

Unadjusted and adjusted competing risk regression analysis for the mortality outcomes according to BMI categories. Each model was adjusted to account for age, sex, CCI and APACHE-II scores. In each analysis, the sub-distribution hazard ratio (SHR) denotes a comparison for each BMI category to the healthy-weight category for cumulative incidence of the outcome of interest.

Obesity groups SHR (95% CI), p value
(Unadjusted)		 SHR (95% CI), p value
(Adjusted)
28-day hospital mortality
Healthy 1.0 1.0
Overweight 1.50 (0.69, 3.26), p = 0.31 2.01 (0.85, 4.76), p = 0.11
BMI category I 1.45 (0.65, 3.23), p = 0.37 3.32 (1.30, 8.46), p = 0.01∗
BMI category II 1.00 (0.39, 2.55), p = 0.99 2.68 (0.89, 8.03), p = 0.08
BMI category III 0.81 (0.27, 2.43), p = 0.70 2.39 (0.73, 7.85), p = 0.15
Overall hospital mortality
Healthy 1.0 1.0
Overweight 1.35 (0.70, 2.58), p = 0.37 1.78 (0.85, 3.75), p = 0.13
BMI category I 1.28 (0.65, 2.51), p = 0.48 2.83 (1.30, 6.15), p = 0.009∗
BMI category II 0.86 (0.38, 1.95), p = 0.73 2.13 (0.80, 5.67), p = 0.13
BMI category III 0.67 (0.25, 1.76), p = 0.41 1.82 (0.62, 5.35), p = 0.28
Open in a new tab

Fig. 4.

Fig. 4

Open in a new tab

Competing risk regression charts. These charts show the probability of an event (28-day hospital mortality or overall hospital mortality) in the presence of a competing other event (hospital discharge). Unadjusted 28-day mortality (A) adjusted 28-day mortality (B), unadjusted overall hospital mortality (C), adjusted overall hospital mortality (D). Adjusted for age, sex, Charlson's comorbidity index and APACHE-II score.

4. Discussion

This study explores the relationship between BMI and clinical outcomes in a selective cohort of patients with severe COVID-19 pneumonia with acute hypoxic respiratory failure admitted to an intensive care unit. Raised BMI was a common finding, with 80% of patients being overweight, 53% obese and a median BMI of 31 (IQR 26,35) kg/m2 in the overall cohort. Our study, principally studying the most severe group of COVID-19 patients, within a predominantly Caucasian population (76.5%), suggests that extreme obesity in severe COVID-19 was associated with younger age, less comorbidity and lower disease severity as quantified by cumulative CCI and APACHE-II scores. However, the obesity group III had more pronounced hypoxemia at presentation. There were no differences in the need for invasive mechanical ventilation or RRT between groups. Despite a trend towards lower unadjusted proportionate mortality among the obesity groups II (17%) and III (15%) compared to the healthy group (21.7%), the adjusted (according to age, sex, APACHE-II and CCI) hospital mortality was higher among all obesity groups. However, statistical significance was only reached for BMI group I (BMI 30–34.9 kg/m2), who had nearly three times greater 28-day and overall hospital mortality than the healthy group. Thus, we did not identify the “obesity paradox” phenomenon in this population.

Others have uncovered varying extents of Obesity Paradox in ICU cohorts. A recent study showed that the moderate obesity cohort had a lower mortality rate (13.8%) than healthy-weight patients, or those with overweight or severe obesity (17.6%, 21.7%, and 50%, respectively) [28]. In addition, this study also showed that those with severe obesity tended to develop severe ARDS more frequently with a higher rate of mechanical ventilation [28]. In contrast, our study demonstrated that moderate obesity (BMI 30–34.9 kg/m2) rather than severe obesity (BMI >40 kg/m2) in the context of COVID-19 was independently associated with increased hospital mortality. The discrepancy in the finding could be due to number of reasons including variations in patient demographics and ICU interventions such as the proportion of patients requiring mechanical ventilation. A similar study in an Italian ICU cohort showed no association between BMI and increased risk of 30-day mortality after similar adjustments. Only severe obesity was found to be associated with an increased risk of mortality in invasively ventilated patients [29]. A New York based study concurs with our findings that BMI and length of stay, and BMI and need for mechanical ventilation, were not significantly associated [30]. However, in contrast, they did not observe a relationship between increased BMI and mortality in hospital, postulating that increased mortality for patients with obesity may be as a result of higher comorbidity burden in these patient groups at a population level. Their cohort of patients was also larger (n = 1337) and while there were some similarities with our study in terms of age and comorbidity distribution, this New York cohort was substantially more ethnically diverse, particularly across higher BMI categories.

Meta-analyses have used myriad approaches to isolating obesity as an independent predictor of COVID disease severity and mortality. In two recent meta-analyses [8,9], use of multi-variable analysis to adjust for predefined cofounding variables including prevalence of known risk factors did not prohibit the finding that the risk of intubation and ventilation was increased in people with obesity, and that mortality was associated with increasing BMI. Our study used APACHE-II scores and CCI, both validated for their ability to predict mortality in critically ill patients [31,32], to provide validated assessment of the comorbidity burden and to mitigate the risk of bias arising from inherent differences between subgroups. However, our sample size was modest compared to some other studies and so our study perhaps lacked the power to demonstrate the same findings as the larger metanalyses. Nonetheless, our study adds weight to the growing body of evidence that obesity is an independent risk factor for morbidity and mortality from COVID-19.

Severe COVID-19 infection primarily causes respiratory failure and advanced age is one of the main risk factors associated with poorer clinical outcomes [33]. Moreover, comorbidities such as diabetes mellitus, chronic respiratory, cardiovascular and kidney diseases and immunosuppression are all negatively associated with clinical outcomes in hospitalised patients [[34], [35], [36], [37]]. Obesity imposes a significant additional burden during COVID-19. Obesity increases the risk of infection, hospitalisation and the severity of disease [17,38,39]. However, ICU studies have identified contradictory findings where some postulate the existence of a concept where obesity appears to be protective. This has been demonstrated in ICU conditions such as acute respiratory distress syndrome, heart failure and cancer [[40], [41], [42]]. However, this is often confounded by several mitigating factors such as age, presence of other comorbidities and number of organ failures present during ICU admission. Our finding is comparable with non-COVID-19 ARDS patients, where the obesity paradox is associated with younger age with lower disease severity [43].

Obesity has become a global health concern with a rising prevalence and increased association with other chronic diseases. It is an independent risk factor for pandemic viral infections [44,45]. Obesity is linked with an increased risk of COVID-19 infection and severity. Multiple dysregulated systems may contribute to this association. In obesity, there is an accumulation of abdominal and thoracic adipose tissue, leading to restricted thoracic expansion and compromised ventilation at the lung bases worsening hypoxaemia [46]. Obesity also disrupts innate and adaptive immune responses [47,48]. Low-grade inflammation associated with obesity can contribute to the recruitment of inflammatory cells and overt inflammatory response contributing to the “cytokine storm” often seen in severe COVID-19 patients [[49], [50], [51]]. Obese patients are also more susceptible to the coagulopathy induced by COVID-19 because adipose tissues produce excess plasminogen activator inhibitor-1 (PAI-1), which impairs fibrinolysis. Chronic inflammation in obesity drives procoagulant factors such as P-selectin and adhesion molecules and inhibits anticoagulant regulatory proteins, such as antithrombin and protein C, which leads to fewer checks on the procoagulant effects of SARS-CoV-2. It has been suggested that ACE2 receptors may be disproportionately present in adipose tissue [52], permitting viral entry via the spike protein and allowing adipose stores to act as a viral reservoir. Obesity is also associated with overexpression of angiotensin II at the systemic level [53], resulting in pulmonary vasoconstriction and enhanced microvascular permeability, both hallmarks of acute lung injury [54,55]. These multiple pathophysiological processes in combination may explain the association between the increased severity of COVID-19 and obesity.

Our study has several limitations. This is from a single centre and has a retrospective observational design. The sample size is modest with a lower mortality rate compared to internationally published studies and inclusive of only few non-obese patients. As population characteristics and ICU interventions may vary between centres, the findings may not be transferable universally. Nevertheless, our study adds to the growing literature on outcomes for obese patients with severe COVID-19 pneumonia requiring respiratory support in the intensive care setting. Our findings are consistent with several studies of similar design that have demonstrated that there are multiple confounding factors which may influence overall patient outcome.

5. Conclusions

The relationship between BMI and morbidity and mortality in our dataset was not straightforward. Our data do not confirm that obesity is paradoxically protective in the highest BMI subgroups with these patients at lower risk of mortality; however, higher BMI patients are more likely to be young, with single organ failure, less chronic comorbidity but with increased severity of hypoxemia at presentation. Once adjusted for these variables, obesity was associated with significant increase in mortality only among obesity group I. We have been cautious not to extend the scope of our conclusions beyond the setting in which our cohort has been recruited. The impact of obesity on indicators of disease severity with requirement for organ support measures including invasive mechanical ventilation and renal replacement therapy was not conclusively demonstrated and will require further scrutiny.

Funding statement

No funding.

Declaration of conflict of interests

Authors declares no conflict of interests.

Detail of authors contribution

Literature search – MJ, AD.

Data collection – MB, LH, EH, CM.

Study design – AD, CM.

Data analysis – AD, HM.

Manuscript preparation – MJ, AD, KS, PA, PC.

Review of manuscript – All authors.

Acknowledgements

We thank all the members of the REACT study group and the GICU consultants' group.

The REACT study group consist of Tom Wilkinson, Anna Freeman, Hannah Burke, Michael Celinski, Saul N Faust, Gareth J Thomas, and Christopher Kipps.

Name and location of the institution where the study was conducted: General Intensive Care Unit, University Hospital Southampton NHS Foundation Trust, UK.

Contributor Information

the REACT COVID-19 Investigators:

Tom Wilkinson, Anna Freeman, Hannah Burke, Ahilanandan Dushianthan, Michael Celinski, Saul N. Faust, Gareth J. Thomas, and Christopher Kipps

References

  • 1.WHO COVID-19 Dashboard . 2020. Geneva world heal organ.https://covid19.who.int/ [Google Scholar]
  • 2.Grasselli G., Zangrillo A., Zanella A., Antonelli M., Cabrini L., Castelli A., et al. Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the lombardy region. Italy. JAMA. 2020;323:1574–1581. doi: 10.1001/jama.2020.5394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Zhou F., Yu T., Du R., Fan G., Liu Y., Liu Z., et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020;395:1054–1062. doi: 10.1016/S0140-6736(20)30566-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Richardson S., Hirsch J.S., Narasimhan M., Crawford J.M., McGinn T., Davidson K.W., et al. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York city area. JAMA. 2020;323:2052–2059. doi: 10.1001/jama.2020.6775. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based measurement studies with 19·2 million participants. Lancet. 2016;387:1377–1396. doi: 10.1016/S0140-6736(16)30054-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Obesity and overweight. WHO.INT; 2021. [Google Scholar]
  • 7.Almond M.H., Edwards M.R., Barclay W.S., Johnston S.L. Obesity and susceptibility to severe outcomes following respiratory viral infection. Thorax. 2013;68:684–686. doi: 10.1136/thoraxjnl-2012-203009. [DOI] [PubMed] [Google Scholar]
  • 8.Chetboun M., Raverdy V., Labreuche J., Simonnet A., Wallet F., Caussy C., et al. BMI and pneumonia outcomes in critically ill covid-19 patients: an international multicenter study. Obesity. 2021;29:1477–1486. doi: 10.1002/oby.23223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Du Y., Lv Y., Zha W., Zhou N., Hong X. Association of body mass index (BMI) with critical COVID-19 and in-hospital mortality: a dose-response meta-analysis. Metabolism. 2021;117 doi: 10.1016/j.metabol.2020.154373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Ho J.S.Y., Fernando D.I., Chan M.Y., Hui Sia C. Obesity in COVID-19: a systematic review and meta-analysis. Ann Acad Med Singapore. 2020;49(12):996–1008. doi: 10.47102/annals-acadmedsg.2020299. [DOI] [PubMed] [Google Scholar]
  • 11.Huang Y., Lu Y., Huang Y.-M., Wang M., Ling W., Sui Y., et al. Obesity in patients with COVID-19: a systematic review and meta-analysis. Metabolism. 2020;113 doi: 10.1016/j.metabol.2020.154378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Kompaniyets L., Goodman A.B., Belay B., Freedman D.S., Sucosky M.S., Lange S.J., et al. Body mass index and risk for COVID-19-related hospitalization, intensive care unit admission, invasive mechanical ventilation, and death - United States, march-december 2020. MMWR Morb Mortal Wkly Rep. 2021;70:355–361. doi: 10.15585/mmwr.mm7010e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Poly T.N., Islam M.M., Yang H.C., Lin M.C., Jian W.-S., Hsu M.-H., et al. Obesity and mortality among patients diagnosed with COVID-19: a systematic review and meta-analysis. Front Med. 2021;8:28. doi: 10.3389/fmed.2021.620044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Deng L., Zhang J., Wang M., Chen L. Obesity is associated with severe COVID-19 but not death: a dose-response meta-analysis. Epidemiol Infect. 2021;149 doi: 10.1017/S0950268820003179. e144–e144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Biscarini S., Colaneri M., Ludovisi S., Seminari E., Pieri T.C., Valsecchi P., et al. The obesity paradox: analysis from the SMAtteo COvid-19 REgistry (SMACORE) cohort. Nutr Metabol Cardiovasc Dis. 2020;30:1920–1925. doi: 10.1016/j.numecd.2020.07.047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Helvaci N., Eyupoglu N.D., Karabulut E., Yildiz B.O. Prevalence of obesity and its impact on outcome in patients with COVID-19: a systematic review and meta-analysis. Front Endocrinol. 2021;12:82. doi: 10.3389/fendo.2021.598249. https://doi:10.3389/fendo.2021.598249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Zhang X., Lewis A.M., Moley J.R., Brestoff J.R. A systematic review and meta-analysis of obesity and COVID-19 outcomes. Sci Rep. 2021;11:7193. doi: 10.1038/s41598-021-86694-1. https://doi:10.1038/s41598-021-86694-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.ICNARC report on COVID-19 in critical care. 2020. https://www.icnarc.org/our-audit/audits/cmp/reports [Google Scholar]
  • 19.Schavemaker R., Schultz M.J., Lagrand W.K., van Slobbe-Bijlsma E.R., Serpa Neto A.P.F. Associations of body mass index with ventilation management and clinical outcomes in invasively ventilated patients with ARDS related to COVID-19—insights from the PRoVENT-COVID study. J Clin Med. 2021;10(6):1176. doi: 10.3390/jcm10061176. https://doi:10.3390/jcm10061176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Naderi N., Kleine C.-E., Park C., Hsiung J.-T., Soohoo M., Tantisattamo E., et al. Obesity paradox in advanced kidney disease: from bedside to the bench. Prog Cardiovasc Dis. 2018;61:168–181. doi: 10.1016/j.pcad.2018.07.001. https://doi:10.1016/j.pcad.2018.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Chittal P., Babu A.S., Lavie C.J. Obesity paradox: does fat alter outcomes in chronic obstructive pulmonary disease? COPD. 2015;12:14–18. doi: 10.3109/15412555.2014.915934. [DOI] [PubMed] [Google Scholar]
  • 22.Keller K., Hobohm L., Münzel T., Ostad M.A., Espinola-Klein C., Lavie C.J., et al. Survival benefit of obese patients with pulmonary embolism. Mayo Clin Proc. 2019;94:1960–1973. doi: 10.1016/j.mayocp.2019.04.035. [DOI] [PubMed] [Google Scholar]
  • 23.Lavie C.J., Coursin D.B., Long M.T. The obesity paradox in infections and implications for COVID-19. Mayo Clin Proc. 2021;96:518–520. doi: 10.1016/j.mayocp.2021.01.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Obesity . Natl Inst Heal Care Excell; 2014. identification, assessment and management. [PubMed] [Google Scholar]
  • 25.Burke H., Freeman A., Dushianthan A., Celinski M., Batchelor J., Phan H., et al. Research evaluation alongside clinical treatment in COVID-19 (REACT COVID-19): an observational and biobanking study. BMJ Open. 2021;11 doi: 10.1136/bmjopen-2020-043012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.STROBE Checklists . 2021. Strobe checklists.https://www.strobe-statement.org/checklists/ (accessed December 17, 2021) [Google Scholar]
  • 27.Charlson Comorbidity Index (CCI) - MDCalc . 2021. MDCalcCom.https://www.mdcalc.com/charlson-comorbidity-index-cci (accessed December 17, 2021) [Google Scholar]
  • 28.Dana R., Bannay A., Bourst P., Ziegler C., Losser M.-R., Gibot S., et al. Obesity and mortality in critically ill COVID-19 patients with respiratory failure. Int J Obes. 2021;45:2028–2037. doi: 10.1038/s41366-021-00872-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Halasz G., Leoni M.L.G., Villani G.Q., Nolli M., Villani M. Obesity, overweight and survival in critically ill patients with SARS-CoV-2 pneumonia: is there an obesity paradox? Preliminary results from Italy. Eur J Prev Cardiol. 2020 doi: 10.1177/2047487320939675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Plataki M., Pan D., Goyal P., Hoffman K., Choi J.M.K., Huang H., et al. Association of body mass index with morbidity in patients hospitalised with COVID-19. BMJ Open Respir Res. 2021;8 doi: 10.1136/bmjresp-2021-000970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Quach S., Hennessy D.A., Faris P., Fong A., Quan H., Doig C. A comparison between the Apache II and Charlson Index Score for predicting hospital mortality in critically ill patients. BMC Health Serv Res. 2009;9:129. doi: 10.1186/1472-6963-9-129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Bouch D.C., Thompson J.P. Severity scoring systems in the critically ill. Cont Educ Anaesth Crit Care Pain. 2008;8:181–185. doi: 10.1093/bjaceaccp/mkn033. [DOI] [Google Scholar]
  • 33.Liu Y., Mao B., Liang S., Yang J.-W., Lu H.-W., Chai Y.-H., et al. Association between age and clinical characteristics and outcomes of COVID-19. Eur Respir J. 2020;55 doi: 10.1183/13993003.01112-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Zhou Y., Yang Q., Chi J., Dong B., Lv W., Shen L., et al. Comorbidities and the risk of severe or fatal outcomes associated with coronavirus disease 2019: a systematic review and meta-analysis. Int J Infect Dis. 2020;99:47–56. doi: 10.1016/j.ijid.2020.07.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Fernández Villalobos N.V., Ott J.J., Klett-Tammen C.J., Bockey A., Vanella P., Krause G., et al. Effect modification of the association between comorbidities and severe course of COVID-19 disease by age of study participants: a systematic review and meta-analysis. Syst Rev. 2021;10:194. doi: 10.1186/s13643-021-01732-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Izzi-Engbeaya C., Distaso W., Amin A., Yang W., Idowu O., Kenkre J.S., et al. Adverse outcomes in COVID-19 and diabetes: a retrospective cohort study from three London teaching hospitals. BMJ Open Diabetes Res & Amp; Care. 2021;9 doi: 10.1136/bmjdrc-2020-001858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Gao Y., Chen Y., Liu M., Shi S., Tian J. Impacts of immunosuppression and immunodeficiency on COVID-19: a systematic review and meta-analysis. J Infect. 2020;81:e93–e95. doi: 10.1016/j.jinf.2020.05.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Popkin B.M., Du S., Green W.D., Beck M.A., Algaith T., Herbst C.H., et al. Individuals with obesity and COVID-19: a global perspective on the epidemiology and biological relationships. Obes Rev. 2020;21 doi: 10.1111/obr.13128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Cai Z., Yang Y., Zhang J. Obesity is associated with severe disease and mortality in patients with coronavirus disease 2019 (COVID-19): a meta-analysis. BMC Publ Health. 2021;21:1505. doi: 10.1186/s12889-021-11546-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Zhi G., Xin W., Ying W., Guohong X., Shuying L. Obesity paradox” in acute respiratory distress syndrome: asystematic review and meta-analysis. PLoS One. 2016;11 doi: 10.1371/journal.pone.0163677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Lavie C.J., Sharma A., Alpert M.A., De Schutter A., Lopez-Jimenez F., Milani R.V., et al. Update on obesity and obesity paradox in heart failure. Prog Cardiovasc Dis. 2016;58:393–400. doi: 10.1016/j.pcad.2015.12.003. [DOI] [PubMed] [Google Scholar]
  • 42.Lee D.H., Giovannucci E.L. The obesity paradox in cancer: epidemiologic insights and perspectives. Curr Nutr Rep. 2019;8:175–181. doi: 10.1007/s13668-019-00280-6. [DOI] [PubMed] [Google Scholar]
  • 43.Ball L., Serpa Neto A., Pelosi P. Obesity and survival in critically ill patients with acute respiratory distress syndrome: a paradox within the paradox. Crit Care. 2017;21:114. doi: 10.1186/s13054-017-1682-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Louie J.K., Acosta M., Samuel M.C., Schechter R., Vugia D.J., Harriman K., et al. A novel risk factor for a novel virus: obesity and 2009 pandemic influenza A (H1N1) Clin Infect Dis. 2011;52:301–312. doi: 10.1093/cid/ciq152. [DOI] [PubMed] [Google Scholar]
  • 45.Halim A.A., Alsayed B., Embarak S., Yaseen T., Dabbous S. Clinical characteristics and outcome of ICU admitted MERS corona virus infected patients. Egypt J Chest Dis Tuberc. 2016;65:81–87. doi: 10.1016/j.ejcdt.2015.11.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Hendren N.S., de Lemos J.A., Ayers C., Das S.R., Rao A., Carter S., et al. Association of body mass index and age with morbidity and mortality in patients hospitalized with COVID-19. Circulation. 2021;143:135–144. doi: 10.1161/CIRCULATIONAHA.120.051936. [DOI] [PubMed] [Google Scholar]
  • 47.Giamarellos-Bourboulis E.J., Netea M.G., Rovina N., Akinosoglou K., Antoniadou A., Antonakos N., et al. Complex immune dysregulation in COVID-19 patients with severe respiratory failure. Cell Host Microbe. 2020;27:992–1000. doi: 10.1016/j.chom.2020.04.009. e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Luzi L., Radaelli M.G. Influenza and obesity: its odd relationship and the lessons for COVID-19 pandemic. Acta Diabetol. 2020;57:759–764. doi: 10.1007/s00592-020-01522-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Tsigalou C., Vallianou N., Dalamaga M. Autoantibody production in obesity: is there evidence for a link between obesity and autoimmunity? Curr Obes Rep. 2020;9:245–254. doi: 10.1007/s13679-020-00397-8. [DOI] [PubMed] [Google Scholar]
  • 50.Korakas E., Ikonomidis I., Kousathana F., Balampanis K., Kountouri A., Raptis A., et al. Obesity and COVID-19: immune and metabolic derangement as a possible link to adverse clinical outcomes. Am J Physiol Endocrinol Metab. 2020;319:E105–E109. doi: 10.1152/ajpendo.00198.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Fajgenbaum D.C., June C.H. Cytokine storm. N Engl J Med. 2020;383:2255–2273. doi: 10.1056/NEJMra2026131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Al-Benna S. Association of high level gene expression of ACE2 in adipose tissue with mortality of COVID-19 infection in obese patients. Obes Med. 2020;19 doi: 10.1016/j.obmed.2020.100283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Iannelli A., Favre G., Frey S., Esnault V., Gugenheim J., Bouam S., et al. Obesity and COVID-19: ACE 2, the missing tile. Obes Surg. 2020;30:4615–4617. doi: 10.1007/s11695-020-04734-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Imai Y., Kuba K., Rao S., Huan Y., Guo F., Guan B., et al. Angiotensin-converting enzyme 2 protects from severe acute lung failure. Nature. 2005;436:112–116. doi: 10.1038/nature03712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Kuba K., Imai Y., Rao S., Gao H., Guo F., Guan B., et al. A crucial role of angiotensin converting enzyme 2 (ACE2) in SARS coronavirus-induced lung injury. Nat Med. 2005;11:875–879. doi: 10.1038/nm1267. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Clinical Nutrition Espen are provided here courtesy of Elsevier

RESOURCES