Discrepancies Between Blood Glucose and Interstitial Glucose: Technological Artifacts or Physiology: A Reply

Sirs,

Differences between blood glucose (BG) and interstitial fluid (ISF) occur when glucose is changing in non–steady state, due to diffusion between compartments (popular notion “lag time”). Because the glucose concentration in the ISF lags behind BG, CGM was not recommended for therapeutic decisions alone for a long time.

Observations of differences between these compartments that do not fit in this classical dynamic concept were recently reported in your journal by Siegmund and colleagues.¹ These are occurring especially under extreme situations like cycling marathon or repetitive hypoglycemia. Regarding their importance for therapeutic decisions, such observations of CGM-experienced clinicians can be of great value for generating new insights into metabolism and/or hypothesis for future studies. However, the authors believe that these differences are physiology based and therefore argue that “therapeutic decisions based on BG values might be inappropriate or even dangerous.”

In my opinion, caution is required for interpreting the causes without sufficient scientific evidence and deriving solutions for therapy planning. The purpose of this reply is (i) to question the physiological interactions, described by the authors, (ii) to refer to studies having investigated the compartmental difference in more detail, and (iii) to offer alternative explanations for the differences.

• (i) In a straightforward approach (Entity 2), the authors assume that glycogen storage plays an essential role, particularly after food ingestion. For a more detailed explanation, they refer to their book:² “If exogenous glucose is offered to the organism, at first this is flowing to areas or organs, where a deficit prevails, if any. These are the brain and empty glycogen stores. This glucose cannot flow into the ISF, where it could be detected by a sensor. Thus, interstitial and plasma glucose are different in such situations.”

• Organs remote from the sensor like the liver and brain, particularly when uptake from tissues equilibrates rapidly with plasma, can affect the BG, but not the plasma/ISF difference in steady state and to a minor extend in dynamic state. To measure signals that are significantly different for therapeutic decisions, the sensor should be placed in the ISF bathing these organs^3,4 (and personal communication from Dr G. Steil, Boston). A gap lasting for some minutes between ISF and plasma could also result from an IV bolus injection of glucose, however, meals or a rapid utilization in tissues are not supposed to produce such a temporary disequilibrium.

• (ii) As can be seen in Figure 1A in Siegmund et al¹ (Entity 1), the glucose concentration is relatively stable; however, differences between ISF and BG are occurring of more than 100 mg/dl. In the past 20 years, several studies^3-6 have been published including sampling from deep and superficial lymphatic system draining muscle under basal and super physiologic insulin levels with and without glucose tracers, clamp-studies and mathematical modeling, mainly a 2-compartment model, where the ISF-to-plasma difference and the time constant τ can be estimated by identification.

• These studies have shown that concentration differences between plasma and ISF are dominated by diffusion; as more glucose is utilized in the surrounding tissue (muscle, fat), so more pronounced is this gap. The ISF-concentration amounts to 85-100% of arterial blood, in one study this was 60% on average.⁶ The big differences shown by the authors have not been found in the studies mentioned above.

• (iii) There are other possible sources that may explain the observations in Siegmund et al.¹ The difference between BG and sensor current is based on three different steps:

• (1) The real difference between plasma and ISF,

• (2) The unknown ratio between sensor current and ISF, and

• (3) The correct estimation of the calibration-factor (personal communication from A. Facchinetti, Padova).⁴

• A calibration procedure, done at rest, is potentially corrupted during or after a marathon because of the sensor-tissue interface (local alteration in microcirculation, tissue e and other effects). Such alterations are a much more plausible explanation for the observed effects (personal communication from W. Regittnig, Graz).

It should be welcomed that diabetologists report data that are obviously not in common with current state of scientific knowledge. This could initialize a broad discussion whether the authors’ interpretation is consistent with physiology, and whether the big differences observed between ISF and BG that disagree with past experimental results might more likely stem from other sources such as stated in (iii). Studies with multiple sensors (from different companies) to rule out a sensing/calibration problem and/or studies with tracers and exercise as planned by Schiavon et al⁵ evaluating the effect of non-steady-state conditions (ie, meals and exercise) should be awaited. In addition, randomized controlled trials, as proposed by the authors, would be valuable. In conclusion the change from BG to CGM as actionable guidance in a therapeutic decision when both differ significantly should be ensured by further studies, particularly before offered to patients in training courses or to doctors in continuing medical education. Hopefully this reply is helpful to initiate such further steps.