Electrochemical Sensors for Measurement of Glucose in Tiny Volumes of Tear Fluid

Tuesday, May 13, 2014: 08:30
Floridian Ballroom L, Lobby Level (Hilton Orlando Bonnet Creek)
M. E. Meyerhoff, K. H. Cha, Y. Qin (University of Michigan), B. C. Cohan (EyeLab Group, LLC), A. Balijepalli, and G. C. Jensen (University of Michigan)
Tear glucose measurements have been suggested as a potential non-invasive alternative to blood glucose monitoring for diabetic patients [1,2].  While previous work has demonstrated a correlation between blood and tear glucose values in humans using complex LC-MS and capillary electrophoresis methods, this correlation has not been thoroughly established and additional clinical studies with a much larger patient population are sorely needed.  This requires the availability of a very simple device that can quantitate the levels of glucose in tiny volumes (1 uL or less) of human tear fluid at the point of testing.  Further, given the low levels of glucose found in tear fluid (10-300 µM), compared to blood plasma (3-15 mM), this creates a challenge for the conventional electrochemical measurement technologies typically employed for blood glucose measurements. 

In preliminary efforts toward achieving this goal, we utilized miniaturized needle type amperometric glucose enzyme electrodes that can be inserted into a microliter glass capillary tube  to measure the levels of glucose in tear fluid obtained from anesthetized rabbits and proved that there was a reasonable correlation between blood and tear glucose levels  [3,4].  However, the sensors employed in these first studies required 3 µL or more of tear fluid and the capillary tube was used to conveniently sample such volumes.   

A more attractive approach for determining glucose in < 1 µL tear volumes from human subjects is to use commercial electrochemical blood glucometer test strips in association with a laboratory quality, high current sensitivity potentiostat.   Of many commercial blood glucose test strips evaluated, only one device has been found to be capable of achieving the low detection limits (< 10 µM) required for human tear glucose measurements.  Using such strips, calibration data in the range of 0-300 μM glucose yield excellent precision and good selectivity over potential redox species found in tears (e.g., ascorbic and uric acids).  Multiple measurements of glucose in tear fluid obtained from normal (non-diabetic) human subjects using commercial strips in conjunction with a  laboratory potentiostat yield fasting tear glucose concentrations for non-diabetics within the range of 20-160 µM, similar to the values found by others using more complex LC-MS methods [1].   This glucometer strip based method could greatly facilitate more in depth clinical studies with a large number of patients (both normal and diabetic), to determine definitively whether tear glucose and blood glucose levels correlate well enough to enable routine measurements in tear fluid to supplement blood glucose testing.


[1]  Lane JD, et al. Curr. Eye Research 31 (2006), 895–901.

[2]  Baca JT, et al. Clin. Chem. 53 (2007), 1370-1372.

[3]  Yan Q, et al. Anal. Chem. 83 (2011), 8341-8346.

[4]  Peng B, et al. Biosens. Bioelectron. (2013) 49, 204-209.