PSUN146 Diabetes, Obesity, and Diabetes-Obesity Overlap Syndrome in Hospitalized Adult COVID-19 Patients: Comorbidity Dose is The Thing!

Abstract

Abstract Excess risk for mortality in hospitalized adults with COVID-19 has been linked to extant chronic conditions, especially obesity (Obes), diabetes (DM) or combination (ObesDM). We investigated dose-response interactions between Elixhauser Comorbidity Index (ECI) and comorbidity ensembles including Obes, DM, ObesDM or Control. Presentation demographics, putative COVID-19 severity markers, and administrative data including ICD-10 codes to measure ECI (AHRQ, v2022.1, 0-38 categories) were extracted under IRB exemption from electronic medical records. Youden's J statistic identified ECI35 prognostic of mortality. Bootstrap Forest (BF) estimated explained variance (EV%) in mortality provided by ECI and comorbidity ensembles including Obes, DM, or ObesDM. Continuous data summarized with median [IQR] were compared using Kruskal-Wallis ANOVA. Discrete data summarized as proportions were compared with chi-squared test. Confounders statistically balanced included age, sex, race, COVID-19 directed treatment and 4-surges of local pandemic. Significant p-value (.013) was Bonferroni corrected. Among 4,275 consecutive COVID-19 patients discharged between March 14, 2020 and September 30, 2021, there were 834 (Obes), 730 (DM), 610 (ObesDM) and 2,101 (Control). Intergroup results are reported using same sequence. Pooled age 69[56-79] years, among 45% females was distributed across Whites (78%), Blacks (10%) and other (12%) races. Additional chronic conditions exhibiting intergroup differences (p<.013) included pooled hypertension (47%, 65%, 58%, 55%), pooled kidney disease (10%, 38%, 25%, 17%), deficiency anemias (19%, 27%, 29%, 23%), chronic pulmonary disease (20%, 19%, 25%, 24%), pooled neurological disorders (14%, 22%, 23%, 23%), heart failure (12%, 21%, 26%, 14%), pooled thyroid disease (16%, 18%, 16%, 17%), coagulopathy (14%, 15%, 14%, 18%) and depression (17%, 16%, 17%, 19%). ECI was 3[2-5], 4[3-6], 5[4-7], vs 3[2-5] (p<.013). ECI accounted for 81% of EV in mortality versus comorbidity ensembles including Obes (11%), ObesDM (4%) or DM (4%). Obes, DM, ObesDM or Control patients with ECI35 vs not exhibited mortality of 25 vs 9% (p<.001), 13 vs 7% (p=.01), 22 vs 7% (p<.001) or 15 vs 6% (p=.01). Statistically equivalent inflammatory markers included CRP (8.2[4.2-13.5], 7.0[2.8-12.5], 9.6[4.8-15.2], 6.3[2.3-11.8] mg/dL, LDH (366[273-488], 322[246-429], 352[264-477], 295[222-392] U/L), ferritin (565[236-1150], 563[255-1166], 549[245-990], 435[200-931] ng/mL) and D-dimer (0.84 [0.53-1.67], 1.04[0.60-1.97], 0.89[0.50-1.78], 1.01[0.56-2.05] mg/mL. Metabolic markers exhibited intergroup differences (p<.013) including cholesterol 151[104-173], 137[104-164], 142[113-179], 143[113-178] mg/dL, triglycerides (161[109-258], 146[100-217], 179[122-252], 120[85-182] mg/dL) and serum glucose 113[101-128], 156[119-229], 158[119-229], 172[126-239] mg/dL. ICU admission was 24%, 19%, 31%, 16% with ObesDM and Obes p<.013 vs DM and control. In conclusion, we found dose of comorbidities in hospitalized COVID-19 patients was more strongly associated with mortality than confounder balanced Obes, DM, or ObesDM. Further investigation is warranted to characterize subgroups with varied response to increasing polymorbidity. Interpretation caveats include monocenter retrospective study with potential unmeasured confounders. Presentation: Sunday, June 12, 2022 12:30 p.m. - 2:30 p.m.