(439) Impact of Obesity and BMI on Glycemic Control in Critically Ill Receiving Parenteral Nutrition

Introduction: Although insulin resistance is more common in patients with obesity, there is limited data on glycemic control in critically ill patients receiving parenteral nutrition (PN) with obesity and across BMI classifications. This study aims to assess the impact of obesity and BMI on glycemic control in critically ill adults initiated on PN.

Methods: In this retrospective cohort study, critically ill adults who received at least two consecutive days of PN from January 2013 to February 2024 were analyzed to assess how obesity (BMI ≥ 30 kg/m2) affects glycemic control. The study compared hyperglycemic episodes (blood glucose > 200 mg/dL), peak blood glucose, and insulin use on day 1 and day 2 of therapy. Glycemic outcomes were compared via Chi-squared, Fisher’s exact, and Wilcoxon rank-sum tests for obesity grouping and Dwass, Steel, Critchlow-Fligner Method across BMI classifications. Multivariate analyses controlled for relevant confounders in assessing effects of obesity and BMI classification on hyperglycemic episodes.

Results: This study included 220 critically ill adults: BMI ≥ 30 kg/m2 (n=91) and BMI < 30 kg/m2 (n=129); BMI classification: underweight: n=18, normal: n=58, overweight: n=53, obesity class I-III: n=91. The overall mean age was 60.3 years and 96.8% received central PN. Although corticosteroid requirements were similar between groups, diabetes mellitus was present in more patients with obesity (38.5% vs 20.2%, p=0.0028). Median dextrose infusion rates (mg/kg/min) were lower for patients with obesity on day 1 (1.2 vs 1.9, p< 0.0001) and day 2 (1.8 vs 2.6, p< 0.0001). Hyperglycemic episodes, peak blood glucose, and insulin use were similar for patients with and without obesity on day 1. On day 2, patients with obesity had a greater number of hyperglycemic episodes (p=0.0478) and insulin use (p=0.0226). Multivariate analyses showed no significant effect of obesity or BMI classification on hyperglycemic episodes.

Conclusions: Despite lower dextrose infusion rates, critically ill adults with obesity experience more hyperglycemic episodes and require more insulin on day 2. However, after controlling for confounders, obesity and BMI classification are not significant predictors of hyperglycemic episodes in the first two days after PN initiation.