Agreement between fingertip-capillary and antecubital-venous appetite-related peptides

This study examined the agreement between fingertip-capillary and antecubital-venous measures of appetite-related peptides. Simultaneous fingertip-capillary and antecubital-venous blood samples were collected from 19 participants. The samples were obtained at baseline, 30, 60, 90, and 120 min following breakfast for the determination of plasma GLP17-36, glucagon, insulin and leptin. Between-day reproducibility of fingertip-capillary-derived estimates was assessed in 18 participants. Deming regression, limits of agreement (LOA) and typical error as a coefficient of variation (CV) were used to quantify agreement (CVa) and reproducibility (CVr). Deming regression revealed no systematic bias for any of the analytes studied, but for insulin there was evidence of a proportional difference at higher concentrations. Measures of GLP17-36 (CVa=24.0%, LOA ±2.5 pg m/l per h), leptin (CVa=9.0%, LOA ×/÷1.19) and glucagon (CVa=21.0%, LOA, ±31.5 pg m/l per h) revealed good agreement between methodological approaches. Fingertip-capillary glucagon was highly reproducible between days (CVr=8.2%). GLP17-36 and leptin demonstrated modest reproducibility (CVr=22.7 and 25.0% respectively). For insulin, agreement (CVa=36.0%, LOA ×/÷1.79) and reproducibility were poor (CVr=36.0%). Collectively, the data demonstrate that fingertip-capillary blood sampling provides a comparable and reproducible alternative to antecubital-venous blood sampling for the quantification of glucagon, and to a lesser extent for GLP17-36 and leptin. Caution should be exercised when utilising fingertip-capillary blood sampling for insulin quantification, and consequently should not be employed interchangeably with antecubital-venous blood sampling.