key: cord-1031477-yaz13ikp
authors: Palaiodimos, L.; Kokkinidis, D. G.; Li, W.; Karamanis, D.; Ognibene, J.; Arora, S.; Southern, W. N.; Mantzoros, C. S.
title: Severe obesity is associated with higher in-hospital mortality in a cohort of patientswith COVID-19 in the Bronx, New York
date: 2020-05-09
journal: nan
DOI: 10.1101/2020.05.05.20091983
sha: 6021ff104a94a0debcdc88c9e921352a0ea3640e
doc_id: 1031477
cord_uid: yaz13ikp

Background & Aims: New York is the current epicenter of Coronavirus disease 2019 (COVID-19) pandemic. The underrepresented minorities, where the prevalence of obesity is higher, appear to be affected disproportionately. Our objectives were to assess the characteristics and early outcomes of patients hospitalized with COVID-19 in the Bronx and investigate whether obesity is associated with worse outcomes. Methods: This retrospective study included the first 200 patients admitted to a tertiary medical center with COVID-19. The electronic medical records were reviewed at least three weeks after admission. The primary endpoint was in-hospital mortality. Results: 200 patients were included (female sex: 102, African American: 102). The median BMI was 30 kg/m2. The median age was 64 years. Hypertension (76%), hyperlipemia (46.2%), and diabetes (39.5%) were the three most common comorbidities. Fever (86%), cough (76.5%), and dyspnea (68%) were the three most common symptoms. 24% died during hospitalization (BMI <25 kg/m2: 31.6%, BMI 25-34 kg/m2: 17.2%, BMI[≥]35 kg/m2: 34.8%, p= 0.03). The multivariate analysis for mortality, demonstrates that BMI[≥]35 kg/m2 (OR: 3.78; 95% CI: 1.45 - 9.83; p=0.006), male sex (OR: 2.74; 95% CI: 1.25 - 5.98; p=0.011) and increasing age (OR: 1.73; 95% CI: 1.13 - 2.63; p=0.011) were independently associated with higher inhospital mortality. Similar results were obtained for the outcomes of increasing oxygen requirement and intubation. Conclusions: In this cohort of hospitalized patients with COVID-19 in a minority-predominant population, severe obesity, increasing age, and male sex were associated with higher in-hospital mortality and in general worse in-hospital outcomes.

confirmed COVID-19 were included. We excluded patients who met one of the following exclusion criteria: i) discharge home directly from the ER, ii) transfer to our center after having received care in other institutions and iii) admission for non-COVID-19 related reasons or non-medical reasons (e.g.

patients admitted because of a fracture, clinically stable patients residing in group homes unable to selfisolate). The 200 included patients were followed for three weeks after their admission to the hospital (admission of 1 st patient: March 9, 2020 ; admission of the 200th patient: March 22, 2020; completion of 3-week follow-up: April 12, 2020) .

Medicine with a waiver of the inform consent (IRB number 2020-11296) .

Two researchers (LP, WL) reviewed the electronic medical records (EMR) independently in a predefined data extraction sheet which was created for the purpose of this study. The documentation of the index admission from emergency medicine providers, inpatient providers, consultants, nurses, therapists, and social workers, the laboratory, and imaging data were reviewed. Post-discharge notes (e.g., tele-medicine follow-up visits, nursing outreach) were also reviewed when available.

Documentation from past visits and the search engine of the EMR were also utilized.

The extracted data included baseline demographic information [(age, gender, race/ethnicity, residence status (community or skilled nursing facility/SNF), and zip code], clinical characteristics [body mass index (BMI), history of smoking, alcohol, intravenous drug use, hypertension, diabetes, hyperlipidemia, coronary artery disease (CAD), heart failure, cerebrovascular disease, chronic obstructive pulmonary . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) disease (COPD), asthma, active malignancy, chronic kidney disease (CKD) or end-stage renal disease (ESRD), liver cirrhosis, and human immunodeficiency virus infection (HIV) or acquired immunodeficiency syndrome (AIDS)], pertinent home medications (immunosuppressive agents, ace-inhibitors, angiotensin II receptor blockers), symptomatology since disease onset and on presentation (fever, headache, malaise, myalgia, rhinorrhea, nasal congestion, sore throat, chest pain, dyspnea, cough, sputum production, nausea/vomiting, diarrhea), vital signs on presentation (oxygen saturation on room air, heart rate, presence of fever), level of oxygen requirement in the ER, laboratory data on the first hospital day [white cell count, lymphocyte count, hemoglobin, platelet count, creatinine, aspartate transaminase (AST), alanine transaminase (ALT), troponin T, creatine kinase (CPK), lactate dehydrogenase (LDH), Ferritin, ddimers, C-reactive protein (CRP), procalcitonin, and hemoglobin A1c for diabetics], initial imaging findings (chest x-ray and/or chest computed tomography), oxygen requirements during hospital stay, acute respiratory distress syndrome (ARDS), intubation, number of days from presentation to intubation, ICU admission, acute kidney injury (AKI) or need for initiation of renal replacement therapy, length of stay, death, and hospital discharge.

The data were processed and analyzed without any personal identifiers to maintain patient confidentiality as per Health Insurance Portability and Accountability Act (HIPAA).

Patients were classified in three groups based on the BMI: BMI<25 kg/m 2 , BMI 25-34 kg/m 2 , and BMI≥35 kg/m 2 as per the most recent BMI assessment prior to or during the index admission. Severe obesity was defined as BMI≥35 kg/m 2 . Patients were also classified in four quartiles based on age: ≤50, 51-64, 65-73, . CC-BY 4.0 International license It is made available under a and ≥74 years old. The primary endpoint was in-hospital mortality. Secondary endpoints included:

increasing oxygen requirement during hospital stay and intubation. Deceased patients were excluded from the length of stay analysis. Continuous data are presented as median with interquartile range (IQR) and categorical data as absolute and relative frequencies. The ANOVA test was used to compare the continuous variables, while chi-square was used for discrete variables. Interaction analyses were performed as needed. A logistic regression model was used to identify baseline variables associated with in-hospital mortality, intubation and increasing oxygen requirements. BMI 25-34 kg/m 2 was used as a reference in order to perform dichotomous comparisons with patients with severe obesity (BMI≥ 35 kg/m 2 ). In order to build a multivariate model, we used a forward stepwise approach with the following method for each one of the studied outcomes; model 1: BMI and age, model 2: all the variables with significant univariate associations (p value ≤ 0.05), and model 3: the variables of model 2 in addition to clinically significant variables which did not show a significant univariate association. Results of logistic regression are given as the odds ratio (OR) with the 95% confidence interval (CI). The threshold of statistical significance was p≤0.05. All analyses were performed using STATA software (version 14·1; STATA Corporation, College Station, TX, USA).

In total, 200 patients admitted with COVID-19 were included in this analysis (female sex= 102, BMI<25 kg/m 2 =38, BMI 25-34 kg/m 2 =116, and BMI≥35 kg/m 2 =46). The median BMI was 30 (IQR 26-35) kg/m 2 .

Most of our patients were either of African American race (51%) or of Hispanic ethnicity (34.5%). 23.5%

were SNF residents. The median age of the whole cohort was 64 (50-73.5) years, with significant differences among the three groups [BMI<25 kg/m 2 : 73 (64-80) vs. BMI 25-34 kg/m 2 63 [48.5-71] vs.

BMI≥35 kg/m 2 : 57.5 (45-67), p<0.001). 32.5% of our cohort was active or past smokers. Hypertension, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 9, 2020 . . https://doi.org/10.1101 /2020 hyperlipemia and coronary artery disease were prevalent in 76%, 46.2% and 16.5% of our patients, respectively. 17% had a history of heart failure while 27.5% had a history of asthma or COPD. 29% percent had a history of chronic kidney disease or ESRD. Diabetes was prevalent in 39.5% of our patients. The detailed baseline demographic and clinical characteristics are presented in Table 1 .

Fever (86%), cough (76.5%), dyspnea (68%) and malaise (58%) were the four most common symptoms.

The median SO 2 on the first hospital day was 95% (IQR 89-97) without significant differences among groups. Symptoms and signs are presented in Table 2 . All 200 patients received chest x-ray on presentation, with 55.5% of them having bilateral infiltrates and 27% having only unilateral findings. The laboratory and radiologic findings are presented in Table 3 .

In total, 24% of our cohort died during hospitalization, with higher rates among individuals with severe obesity (BMI<25 kg/m 2 : 31.6%, BMI 25-34 kg/m 2 : 17.2%, BMI≥35 kg/m 2 : 34.8%, p= 0.030). Similarly, patients with severe obesity were more likely to undergo intubation (BMI<25 kg/m 2 : 18.4%, BMI 25-34 kg/m 2 : 16.4%, BMI≥35 kg/m 2 : 34.8%, p= 0.032). In total, 45% of our patients had increasing oxygen requirements during hospital stay without significant differences among BMI groups. Twenty-two percent developed ARDS and 16% spent at least one night in the ICU. In-hospital outcomes are presented in Table 4 .

The univariate associations with in-hospital mortality were examined for all the baseline demographic and clinical characteristics. Increasing age (analyzed in quartiles), male sex, BMI≥35 kg/m 2 (reference: BMI 25-34 kg/m 2 ), heart failure, CAD, and CKD or ESRD were found to have a significant univariate . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020 . . https://doi.org/10.1101 /2020 association ( Table 5 ). The following variables were not shown to be statistically significant in the univariate associations: hypertension, hyperlipidemia, obstructive sleep apnea, diabetes, and smoking ( 1.45 -9.83; p=0.006) were found to have significant associations.

Male sex, current or former smoking and BMI≥35 kg/m 2 (reference: BMI 25-34 kg/m 2 ) were found to have a significant univariate association with increasing oxygen requirements ( (Table 6) .

Male sex, and BMI≥35 kg/m 2 (reference: BMI 25-34.9 kg/m 2 ) were found to have a significant univariate association with intubation (Table 7 ). In the multivariate analysis (model 3), male sex (OR: 3.35; 95% CI:

1.51 -7.46, p=0.003), increasing age analyzed in quartiles (OR: 1.50; 95% CI:1.05 -2.12; p=0.025), and BMI≥35 kg/m 2 (OR: 3.87; 95% CI: 1.47 -10.18; p=0.006) were significant predictors (Table 7) .

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020 . . https://doi.org/10.1101 /2020 Given the significant associations that we noticed for male sex, age and BMI≥35 kg/m 2 with the outcomes that we examined, we performed an interaction analysis for this set of variables. Two different interactions were tested; sex with BMI and age with BMI. Both of them were not significant.

Our study described the baseline characteristics, clinical features, and early outcomes of the first 200 patients who were hospitalized due to COVID-19 in our institution that mainly serves African American and Hispanic population. This is the first study that has performed joint evaluation of age, gender, obesity and multiple comorbidities that have previously been linked with adverse outcomes. The main findings can be summarized as following: 1) the in-hospital mortality was 24% with only 3% patients still hospitalized on the 21-day follow-up 2) severe obesity (BMI≥35 kg/m 2 ), increasing age, and male sex have an independent association with mortality and need for intubation 3) severe obesity (BMI≥35 kg/m 2 ), increasing age, male sex, and smoking have an independent association with increasing oxygen requirements during hospitalization.

Older age and male sex have already been described as risks factor for severe disease and death in patients with [11] [12] [13] [14] [15] [16] , although large outcome studies are needed to assess the latter. The most interesting finding of our analysis is that severe obesity is a significant factor for severe respiratory disease and death in hospitalized patients with COVID-19. It should be pointed out that this association remained significant after adjusting for several clinical entities, such as diabetes, coronary artery disease, heart failure, COPD, CKD or ESRD, and smoking, which indicates that obesity may predispose to negative outcomes independently. To the best of our knowledge, no other studies to date have shown . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. an independent and direct association of severe obesity to mortality in patients with COVID-19; however, some preliminary published data have linked obesity to severe COVID-19. A large cohort from New York City depicted that obesity is strongly associated with progression to critical illness with substantially higher odds ratio than any cardiovascular or pulmonary disease (BMI 30-40 kg/m 2 OR: 1.38; 95% CI: 1.03-1.85; BMI>40 kg/m 2 OR: 1.73; 95% CI 1.03-2.90) [17] . A cohort from China revealed that obesity significantly increases the risk for developing severe pneumonia in the setting of COVID-19 (OR:

3.42; 95% CI: 1.42-8.27) [18] . Another report from China indicated that the presence of obesity in patients with metabolic-associated fatty liver disease is associated with an almost 6-fold increased risk of severe . Obesity could partially explain why the mortality rate for COVID-19 is higher in countries with higher prevalence of obesity, such as Italy, as compared to China and Japan [20] . Variables such as diabetes and other cardiovascular and pulmonary comorbidities were found to have a significant association in prior reports from China and Italy [11] [12] [13] [14] . However, most of these were obtained from univariate estimates only. In our multivariate analyses it was shown that these comorbidities are likely epiphenomena since they are not independent from male gender, older age or obesity which may be the underlying link.

Our findings on the association of severe obesity to mortality in COVID-19 are not unanticipated given our prior experience from the 2009 pandemic influenza (H1N1) disease. Morgan et al. reported significant association of obesity to death in adult patients without recognized pre-existing medical conditions hospitalized with H1N1 influenza disease (obesity OR: 3.1; 95% CI: 1.5 -6.6); morbid obesity OR: 7.6; 95% CI: 2.1 -27.9) [21] . Obesity leads to increased work of breathing by augmenting the airway resistance and is associated with decreased expiratory reserve volume, functional capacity, and pulmonary compliance [22, 23] . Central obesity results in decreased diaphragmatic excursion in supine patients compromising ventilation [23] . Moreover, obesity is a chronic inflammatory state with increased . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. . circulating levels of pro-inflammatory cytokines, including interleukin-6, and is known to impair the immune system [24, 25] . The observed association between severe obesity and mortality may also underlie the observed association between low vitamin D levels, which are low in the obese, and mortality and needs to be explored further [25] .

The in-hospital mortality rate on 21-day follow-up in our cohort is 24%. Three inpatient cohorts from China reported in-hospital mortality rates of 28.2%, 11.7%, and 17%, respectively [5,11,12] . A large cross-sectional study from New York City, which did not include hospitals located in the Bronx, reported a 14.6% mortality rate in the inpatient sample to the time of the analysis, however 35.9% of the patients were still hospitalized, which indicates that the final inpatient mortality rate may be actually higher [16] . Large retrospective cohorts that will probably be published in the following months will accurately estimate the in-hospital mortality in the Bronx and elsewhere.

In total, 33.6% of adults in the Bronx are obese, a number which is much higher than all other NYC boroughs, while the prevalence of obesity nationally is 20% [9, 10] . Additionally, the prevalence of obesity is higher among individuals of lower socioeconomic status (>35%) and substantially higher in non-Hispanic blacks (36%) and Hispanic (35.4%) as compared to non-Hispanic whites (19.1%) [9] . The patient population that our institution serves mainly includes patients of lower socioeconomic status and people that identify themselves as African American or Hispanic. The combination of these two made up 85.5% of our current study population. The median income, as estimated by the zip codes based on publicly available data provided by the internal revenue service, was not found to be a risk factor for worse outcomes in our study, which is explained by the fact that the Bronx is a homogenous area of low income.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. . We report that active or prior smoking was associated with increasing oxygen requirements during hospitalization. This is in contrast to the results of a recent meta-analysis of five studies from China which concluded that active smoking did not seem to be significantly associated with higher risk of progressing to severe . Large observational studies and future meta-analyses will elucidate the association of smoking with COVID-19 severity and other outcomes.

To the best of our knowledge, this is the first study to date showing the independent association of severe obesity to in-hospital mortality in patients with COVID-19. One of the strengths of the current study is that our included patients represent underserved and economically disadvantaged minorities; thus, revealing the early outcomes of COVID-19 in this vulnerable population and usually underreported and underrepresented in clinical research. Additionally, two researchers independently and blindly collected data which reduces errors and bias. On the other hand, our study has several limitations. First, our sample was relatively small, but given the nature of the evolving pandemic, it was of paramount importance to make our early data and findings widely available as soon as possible, especially given the lack of data up to date in COVID-19 in minorities and underserved population. Second, this was a realworld study with a retrospective design utilizing the electronic medical records, which is suboptimal compared to a prospective study that could have more accurate follow-up assessment. Third, the rapidly changing management of COVID-19 might have affected our results but it highly unlikely that could have differentials affected associations between obesity and mortality. Fourth, we handled BMI as a categorical variable in the regression analysis. This can lead to suboptimal conclusions, but we think that specific cutoffs, following established clinical guidelines on obesity, may be of more interest and ease for the clinicians compared to interpretation of continuous variables in a regression model.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. . In conclusion in this early cohort of hospitalized patients with COVID-19 in an underserved, minoritypredominant population in the Bronx, we found that severe obesity was associated with higher inhospital mortality even after adjusting for other pertinent potential confounding factors. Particular attention should be paid in protection of this population given the higher chance for negative outcomes once they are diagnosed with the disease. In addition, obese patients diagnosed with COVID-19 should be treated with extra caution given the possible higher risk for adverse outcomes. While we recognize the limitations, we hope that our study will stimulate additional researchers to further study the effect of obesity in COVID-19 and outcomes of minorities diagnosed with COVID-19. Larger cohort studies are needed to confirm our data and pilot clinical trials are needed to assess whether pharmacotherapy for obesity and its comorbidities may improve outcomes in the short or the long term. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. . https://doi.org /10.1101 /10. /2020 https://www1.nyc.gov/site/doh/covid/covid-19-data.page . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. (7):438-46.

[23]. Dietz W, Santos-Burgoa C. Obesity and its Implications for COVID-19 Mortality. Obesity. 2020 Apr 1.

[24]. de Heredia FP, Gómez-Martínez S, Marcos A. Obesity, inflammation and the immune system. Proceedings of the Nutrition Society. 2012 May;71 (2):332-8.

[25] Muscogiuri G, Pugliese G, Barrea L, Savastano S, Colao A. Obesity: the "Achilles heel" for COVID-19? Metabolism-Clinical and Experimental. 2020 Apr 27.

[26] Lippi G, Henry BM. Active smoking is not associated with severity of coronavirus disease 2019 . European journal of internal medicine. 2020 Mar 16.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 9, 2020. . Length of stay median (IQR) -days 6 (4-10) 6 (4-12) 5 (4-10) 6 (4-9) 0.924 5 (4-7) 7 (5-12) 6 (4-11.5) 6 (4-11) 0.273

Notes: (1) The outcomes are presented as no. (%) unless specified differently. (2) p-values refer to Chi-square test/ANOVA and stars denote the number of statistically significant pairwise comparisons using Bonferroni's method Abbreviations and symbols: BMI=body mass index, IQR=interquartile range, no.=number, O2=oxygen, ↑=increasing, ARDS=acute respiratory distress syndrome, ICU=intensive care unit, AKI=acute kidney injury, RRT=renal replacement therapy . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 9, 2020 . . https://doi.org/10.1101 /2020 

World Health Organization. Coronavirus disease 2019 (COVID-19) situation report -87

Covid-19 in critically ill patients in the Seattle region-case series

Coronavirus disease 2019 (COVID-19)

Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy region

COVID-19 and African Americans

County Health Rankings Key Findings Report

Montefiore's Office of Community & Population Health Bronx Community Health Dashboard: Nutrition, Physical Activity and Obesity

Adult Obesity Prevalence Maps. Division of Nutrition, Physical Activity, and Obesity, National Center for Chronic Disease Prevention and Health Promotion

Predictors of Mortality for Patients with COVID-19 Pneumonia Caused by SARS-CoV-2: A Prospective Cohort Study

Multivariate Analysis with addition of smoking and diabetes clinical important ones as regressors. Median income was estimated by zip codes. Age in years

Abbreviations: BMI=body mass index, CKD=chronic kidney disease, ESRD=end-stage renal disease, COPD=chronic obstructive pulmonary disease, SNF=skilled nursing facility, ACEi=angiotensin-converting-enzyme inhibitor, ARB=angiotensin II receptor blocker, CI=confidence interval

ACEI or ARB use prior to admission 0

On immunosuppressive therapy 0

Multivariate Analysis with addition of smoking and diabetes clinical important ones as regressors. Median income was estimated by zip codes. Age in years

ESRD=end-stage renal disease, COPD=chronic obstructive pulmonary disease, SNF=skilled nursing facility, ACEi=angiotensin-converting-enzyme inhibitor, ARB=angiotensin II receptor blocker, CI=confidence interval, Ref.=reference Male sex 2.51 (1.23 -5.14)

Residence status (community vs SNF)

Multivariate Analysis with addition of smoking and diabetes clinical important ones as regressors. Median income was estimated by zip codes. Age in years

Abbreviations: BMI=body mass index, CKD=chronic kidney disease, ESRD=end-stage renal disease, COPD=chronic obstructive pulmonary disease, SNF=skilled nursing facility, ACEi=angiotensin-converting-enzyme inhibitor, ARB=angiotensin II receptor blocker, CI=confidence interval

All authors contributed to the manuscript for important intellectual contents and approved the submission.

Conflict of interest statement: All authors declare no conflict of interests