Serial bioelectrical impedance changes during nutritional follow-up in chronic pancreatitis

Nutritional impairment in chronic pancreatitis (CP) may be present even when body mass index (BMI) and routine screening scores do not indicate overt malnutrition. We aimed to provide a bioelectrical impedance analysis (BIA)-based snapshot of body composition in a clinically selected CP cohort referred for nutritional assessment and to evaluate serial changes during follow-up.

This monocentric retrospective analysis included prospectively collected data from the BIOPANCREAS registry (NCT06552078). Adults with probable or definitive CP (M-ANNHEIM criteria) undergoing clinically indicated nutritional assessment including BIA were included. Longitudinal changes in BIA-derived parameters were assessed using linear mixed-effects models with Benjamini-Hochberg false discovery rate (FDR) correction. Complete-case Friedman analyses were performed as sensitivity analyses.

Thirty-two patients were included (mean age 58.4 ± 15.8 years; 65.6% male). At baseline, mean BMI was 22.3 ± 3.8 kg/m² and 77.4% were at low nutritional risk according to the Malnutrition Universal Screening Tool (MUST). Low fat-free mass index (FFMI) and appendicular skeletal muscle mass index (ASMMI) were identified in 18.8% and 40.6%, respectively. BIA measurements were available in 32, 25, and 17 patients at visits I, II, and III. After FDR correction, phase angle (PhA), body cell mass percentage (BCM%), and extracellular water percentage (ECW%) changed significantly. Compared with visit I, PhA increased by 0.24° at visit II (95% confidence interval [CI] 0.06-0.42) and 0.28° at visit III (95% CI 0.07-0.49); BCM% increased by 1.48 (95% CI 0.40-2.56) and 1.80 percentage points (95% CI 0.54-3.05), respectively; and ECW% decreased by 1.41 (95% CI -2.42 to -0.40) and 1.66 percentage points (95% CI -2.84 to -0.49), respectively; All six contrasts remained significant after FDR correction (adjusted p=0.016). No significant visit II to III differences or longitudinal changes in conventional muscle-quantity indices were observed.

In this real-world cohort, BIA identified heterogeneous body-composition abnormalities despite relatively reassuring BMI and MUST findings. During nutritional follow-up, PhA and BCM% increased and ECW% decreased, while conventional muscle-quantity indices did not show significant changes. These hypothesis-generating findings support prospective controlled evaluation of serial BIA during nutritional care in CP.