Continuous noninvasive glucose monitoring; water as a relevant marker of glucose uptake in vivo

Abstract

With diabetes set to become the number 3 killer in the Western hemisphere and proportionally growing in other parts of the world, the subject of noninvasive monitoring of glucose dynamics in blood remains a "hot" topic, with the involvement of many groups worldwide. There is a plethora of techniques involved in this academic push, but the so-called multisensor system with an impedance-based core seems to feature increasingly strongly. However, the symmetrical structure of the glucose molecule and its shielding by the smaller dipoles of water would suggest that this option should be less enticing. Yet there is enough phenomenological evidence to suggest that impedance-based methods are truly sensitive to the biophysical effects of glucose variations in the blood. We have been trying to answer this very fundamental conundrum: "Why is impedance or dielectric spectroscopy sensitive to glucose concentration changes in the blood and why can this be done over a very broad frequency band, including microwaves?" The vistas for medical diagnostics are very enticing. There have been a significant number of papers published that look seriously at this problem. In this review, we want to summarize this body of research and the underlying mechanisms and propose a perspective toward utilizing the phenomena. It is our impression that the current world view on the dielectric response of glucose in solution, as outlined below, will support the further evolution and implementation toward practical noninvasive glucose monitoring solutions.

Keywords: Cole–Cole broadening; Dielectric spectroscopy, Impedance spectroscopy; Glucose monitoring; Microwave; Multisensor; Water.

Fig. 1

Schematic view of the dielectric permittivity (solid line) and loss (dashed) spectrum of a typical biomolecular aqueous solution (Raicu and Feldman ; Schwan 1957). Typically, four relaxations can be noted. Using the nomenclature of the figure, these are as follows: α is related to ionic mobilities, β is related to interfacial polarizations of cellular structures, and γ is the iconic relaxation of bulk water. The source of the intermediate δ is still a point for scientific debate

Fig. 2

The human red blood cell (erythrocyte). The biconcave shape of the cell leads to an asymmetric dielectric response in the β dispersion

Fig. 3

Model for GLUT1-mediated transport (adapted from Heard et al. 2000)). The figure to the left shows a substrate transported through GLUT1. In the absence of ATP or when AMP is bound to GLUT1 (left), sugar can cross easily into the cell. Upon ATP binding (right), a conformational change occurs that restricts sugar transport

Fig. 4

The normalized cell membrane capacitance of biconcave erythrocyte suspensions versus d- (open) and l-glucose (full) concentration. C_m was normalized to 0 mM data because of physiological differences between donors. Mean values ± SD (n = 12). Reproduced with permission from Livshits et al. (2007) (Copyright 2007, IOP)

Fig. 5

Equivalent circuit of an initial fringing field capacitive sensor coupled to skin. In this simplified model, R represents the surface resistance of the skin and C the capacitance of the skin under the electrodes

Fig. 6

Example for one trial day of the normalized (initial value equal 1) ε and σ of the dermis at 30 MHz (left axis) compared to the reference glucose profile (red, x, right axis) versus time. Black +: calculated permittivity (conductivity) of the dermis. Reproduced with permission from Dewarrat et al. (2011). The results show a dependence of the skin permittivity on the glucose concentration in the blood. However, the results obtained per subject were heavily dependent on the state of the skin, the subject, blood perfusion in the skin layer, and other factors

Fig. 7

The value of the capacitor charge at 2 ns (~ 500 MHz) and 3 ns (~ 330 MHz) versus time of the glucose protocol for a single subject. The results were obtained by time domain dielectric spectroscopy, using an open-ended coaxial probe attached to the subject’s skin. The behavior of charge across the probe face mirrors the induced blood glucose concentration

Fig. 8

The complex dielectric permittivity ε(ν), logarithmically displayed as a function of frequency, ν, for a suspension of erythrocytes in an iso-osmolar NaCl aqueous solution (filled circles, volume fraction of RBC = 0.95) and for pure water (open circles), at 25 °C. The dielectric permittivity is presented in terms of its real part, ε′(ν) and the dielectric losses ε״(ν) to the negative imaginary part. The latter quantity has been derived from the total loss, ε” = ε”(ν) − σ_dc/2νε₀, where σ_dc denotes the specific electric dc conductivity. Reproduced with permission from Kaatze (1990) (Copyright 1990, IOP)

Fig. 9

The real part ε′(ν) and the imaginary ε″(ν) of the water dielectric spectra at 25 °C (1021 frequency points) (Levy et al. 2012b) (reproduced with permission from Copyright 2012, American Institute of Physics)

Fig. 10

The four hyperbolic branches of the function defined by Eq. (6) (Puzenko et al. 2010) (reproduced with permission from Copyright 2010, American Institute of Physics)

Fig. 11

3D trajectories of CC relaxation processes of different aqueous solutions at 25 °C: a glucose, b NaCl, c ATP, and d arginine (reproduced with permission from Copyright 2012, 2014 American Institute of Physics). The vertical axis is α, the stretching parameter defined by Eq. (2); the horizontal axis is the logarithm of the relaxation time, τ, defined by the same Eq. (2); and the logarithm of B, the Froehlich B function, defined in Eq. (7). Together these parameters represent the evolution of the dielectric response of bulk water in these solutions as a function of the change of concentration of the solute. In the case of case of glucose (a) and NaCl (b), the relaxation times change monotonically with concentration, depending on whether the interaction of the solute with the solvent (water) is dipole–dipole or dipole–ionic, respectively. In the case of biological molecules such as ATP (c) and arginine (d), the concentration dependence reveals kink points in the dielectric parameters as the dominant behavior varies between dipole–dipole and dipole–ionic

Fig. 12

In panel a are shown the dielectric loss spectra at 25 °C of an RBC suspension in PBS (red triangles) compared to water (black line). Dc conductivity has been removed. In panel b are shown the derived losses of cellular cytoplasm without d-glucose (red triangles) and 10 mM d-glucose (blue circles). The lines are the fitting curves (reproduced with permission from Copyright 2016, American Chemical Society)

Fig. 13

In panel a are shown the experimental relaxation times of cytoplasmic water for one cell in the presence of varying buffer concentrations of d-glucose. In panel b, α(lnτ) is shown dependence for one cell at 25 °C. Solid lines are the fitting curves using Eq. (6) (Levy et al. 2016) (reproduced with permission from Copyright 2016, American Chemical Society)

Fig. 14

The d-glucose concentration dependence of the B function (Levy et al. 2016) as defined by Eq. (7), of the cytoplasmic water relaxation for the red blood cell (reproduced with permission from Copyright 2016, American Chemical Society)

Fig. 15

The elongation ratio of red blood cells in the presence of varying concentrations of d-glucose (blue triangles).The red dots are the relaxation times of the main peak of cytoplasmic water (Levy et al. 2016) for the same concentrations. The critical point, 10 mM, is represented in both data sets and indicates the cessation of glucose uptake by the cell because of the formation of ATP–GLUT1 complexes, effectively blocking the GLUT1 transporter (reproduced with permission from Copyright 2016, American Chemical Society)