Use of c-peptide as a measure of cephalic phase insulin release in humans

Abstract

Cephalic phase insulin release (CPIR) is a rapid pulse of insulin secreted within minutes of food-related sensory stimulation. Understanding the mechanisms underlying CPIR in humans has been hindered by its small observed effect size and high variability within and between studies. One contributing factor to these limitations may be the use of peripherally measured insulin as an indicator of secreted insulin, since a substantial portion of insulin is metabolized by the liver before delivery to peripheral circulation. Here, we investigated the use of c-peptide, which is co-secreted in equimolar amounts to insulin from pancreatic beta cells, as a proxy for insulin secretion during the cephalic phase period. Changes in insulin and c-peptide were monitored in 18 adults over two repeated sessions following oral stimulation with a sucrose-containing gelatin stimulus. We found that, on average, insulin and c-peptide release followed a similar time course over the cephalic phase period, but that c-peptide showed a greater effect size. Importantly, when insulin and c-peptide concentrations were compared across sessions, we found that changes in c-peptide were significantly correlated at the 2 min (r = 0.50, p = 0.03) and 4 min (r = 0.65, p = 0.003) time points, as well as when participants' highest c-peptide concentrations were considered (r = 0.64, p = 0.004). In contrast, no significant correlations were observed for changes in insulin measured from the sessions (r = -0.06-0.35, p > 0.05). Herein, we detail the individual variability of insulin and c-peptide concentrations measured during the cephalic phase period, and identify c-peptide as a valuable metric for insulin secretion alongside insulin concentrations when investigating CPIR.

Keywords: C-peptide; Cephalic phase; Individual differences; Oral stimulation; Preabsorptive insulin release.

Fig. 1.

Overview of the blood draw protocol. Participants received the model gelatin stimulus at the 0 min time point and performed modified sham-feeding for 45 s. Arrows labeled with numbers represent blood collection time points (in minutes) during the baseline (gray) and CPIR (black) periods. Each participant underwent this protocol on two separate days.

Fig. 2.

Changes in insulin (top graphs) and c-peptide (bottom graphs) from the averaged baseline value (labeled BL on the x-axis, set to 0). Data are depicted for session 1 (left panel) and session 2 (right panel). ^†,* indicates a statistically significant difference from zero at p < 0.05 and at an adjusted alpha level of 0.05 after Bonferroni correction (p < 0.0125), respectively.

Fig. 3.

Change in insulin (top graphs) and c-peptide (bottom graphs) concentrations from baseline at session 1 (left panel) and session 2 (right panel). The circles at each time point represent the Δ insulin or Δ c-peptide concentration calculated for each participant (N = 18). Black circles indicate that the concentration was the participant’s highest concentration reached for the marker/session. These points are grouped together on the right of their corresponding graph (labeled “Highest Δ” on the x-axis). A total of 18 circles are shown at each time point (2, 4, 6, 8 min) and the “Highest Δ” on the right. Solid black bars at each time point and on the highest Δ column represent the mean of the grouping. The dotted line at zero represents the mean of the three baseline concentrations (−15, −10, and −5 min).

Fig. 4.

Correlations between Δ c-peptide (x-axis) and Δ insulin (y-axis) concentrations at 2 min (left panel), 4 min (center panel), and highest Δ concentration reached within 2 to 8 min (right panel). The dashed lines represent the 95% confidence interval of the regression line (solid line). Correlation coefficients (r) and corresponding p values are indicated on the top left of each graph. Significant coefficients and p-values are bolded.

Fig. 5.

Correlations between Δ insulin concentrations (top graphs) and Δ c-peptide concentrations (bottom graphs) from two sessions (x-axis: session 1, y-axis: session 2). The figure depicts observations at 2 min (left panel), 4 min (center panel), and highest Δ concentration measured within 2 to 8 min (right panel). The dashed lines represent the 95% confidence interval of the regression line (solid line). Correlation coefficients (r) and corresponding p values are indicated on the top left of each graph. Significant coefficients and p-values are bolded.