Blood glucose amplitude variation: effects of intensive insulin therapy and relative association with mortality

Introduction

There is growing evidence that not only blood glucose (BG) level but also BG amplitude variation (BGAV) is associated with mortality in critically ill patients [1].

Methods

Retrospective analysis of the data of the Leuven intensive insulin therapy (IIT) trial in 1,200 MICU patients, randomized to receive either IIT or conventional insulin therapy [2]. The hyperglycemic index (HGI) and hypoglycemic index (HoGI) were used as measures of BG level, the standard deviation of all BG readings per patient (SD BG) as a measure of BGAV. The univariable effect of IIT on these indices was analyzed, the independent association with hospital mortality was assessed by multivariable logistic regression (MVR), corrected for baseline risks.

Results

IIT reduced the median HGI from 3.2 to 0.8 mmol/l (P < 0.0001), increased the median HoGI from 0.005 to 0.048 mmol/l (P < 0.0001), and did not affect median SD BG (conventional: 2.12; IIT: 1.99 mmol/l (P = 0.161)). The results of the MVR are summarized in Table 1. HGI, HoGI and SD BG were independently associated with mortality.

Table 1

Conclusions

BGAV was associated with mortality in MICU patients, independent of baseline risks and BG level. IIT reduced HGI, increased HoGI, and did not affect BGAV. Reducing BGAV, in addition to IIT, may theoretically increase its potential for clinical benefit.