Plasma FGF-19 Levels are Increased in Patients with Post-Bariatric Hypoglycemia

Abstract

Background: Hypoglycemia is an increasingly recognized complication of bariatric surgery. Mechanisms contributing to glucose lowering remain incompletely understood. We aimed to identify differentially abundant plasma proteins in patients with post-bariatric hypoglycemia (PBH) after Roux-en-Y gastric bypass (RYGB), compared to asymptomatic post-RYGB.

Methods: Proteomic analysis of blood samples collected after overnight fast and mixed meal challenge in individuals with PBH, asymptomatic RYGB, severe obesity, or overweight recruited from outpatient hypoglycemia or bariatric clinics.

Results: The top-ranking differentially abundant protein at 120 min after mixed meal was fibroblast growth factor 19 (FGF-19), an intestinally derived hormone regulated by bile acid-FXR signaling; levels were 2.4-fold higher in PBH vs. asymptomatic post-RYGB (mean + SEM, 1094 ± 141 vs. 428 ± 45, P < 0.001, FDR < 0.01). FGF-19 ELISA confirmed 3.5-fold higher concentrations in PBH versus asymptomatic (360 ± 70 vs. 103 ± 18, P = 0.025). To explore potential links between increased FGF-19 and GLP-1, residual samples from other human studies in which GLP-1 was modulated were assayed. FGF-19 levels did not change in response to infusion of GLP-1 and PYY in overweight/obese individuals. Infusion of the GLP-1 receptor antagonist exendin 9-39 in recently operated asymptomatic post-RYGB did not alter FGF-19 levels after mixed meal. By contrast, GLP-1 receptor antagonist infusion yielded a significant increase in FGF-19 levels after oral glucose in individuals with PBH. While plasma bile acids did not differ between PBH and asymptomatic post-RYGB, these data suggest unique interrelationships between GLP-1 and FGF-19 in PBH.

Conclusions: Taken together, these data support FGF-19 as a potential contributor to insulin-independent pathways driving postprandial hypoglycemia in PBH.

Keywords: Bile acids; FGF-19; Gastric bypass; Hypoglycemia.

Figure 1.

Volcano plot of plasma proteomic data derived from blood samples collected at 120 minutes following mixed meal tolerance test. X-axis present log₂ of ratio of protein content in PBH vs. asymptomatic post-RYGB, with proteins upregulated in PBH plotted to the right, and proteins downregulated plotted to the left; y axis indicates −log₁₀ P value.

Figure 2.

A. Plasma FGF-19 levels in overweight (clear triangle), severely obese (black triangle), asymptomatic post-RYGB (clear square) and PBH (black circle), as measured by SomaLogic platform. Note that SomaLogic analysis was not performed at 120 minutes in overweight or severely obese groups as sample volume was not sufficient for analysis. At time 0, 30 and 120 minutes, P values for comparison between asymptomatic post-RYGB and PBH were 0.02, 0.01 and 6.88E-08, respectively. B. ELISA analysis was performed in samples collected at 120 minutes after mixed meal. * P < 0.05, ** P ≤ 0.01.

Figure 3.

FGF-19 correlates positively with insulin in asymptomatic post-RYGB (clear squares) and PBH (black circle). Insulin levels at 30 minutes (x- axis) and FGF-19 levels (y-axis) are strongly correlated in both bypass groups (r = 0.59, P = 0.01) (black solid line). The correlation coefficient is higher in participants with neuroglycopenia (r = 0.71, P = 0.02) (black dotted line).

Figure 4.

Plasma FGF-19 levels prior to and following a 75 gram oral glucose load (ingested in portions over twenty minutes, marked by shaded area) in participants with PBH and neuroglycopenia. Participants received blinded infusion of either exendin-9–39 (black circles) or placebo (saline) (clear squares) in a crossover fashion. * p < 0.05, *** p <0.0001 for comparison of placebo vs. exendin 9–39. ANOVA for time, treatment and the interaction of time and treatment p < 0.001, p = 0.02 and p=0.0507, respectively.