Performance of Glucose Dehydrogenase–and Glucose Oxidase–Based Blood Glucose Meters at High Altitude and Low Temperature
- Daniel Öberg, MD and
- Claes-Göran Östenson, MD, PHD
- From the Department of Molecular Medicine, Endocrine and Diabetes Unit, Karolinska University Hospital, Stockholm, Sweden
- Address correspondence to Daniel Öberg, c/o Claes-Göran Östenson, Department of Molecular Medicine, Endocrine and Diabetes Unit, Karolinska University Hospital, 171 76 Stockholm, Sweden. E-mail: daniel{at}oberg.se
Blood glucose meters using the enzyme glucose oxidase (GO) have been proven unreliable at high altitude (1–6). A new test strip technology, based on the oxygen-insensitive enzyme glucose dehydrogenase (GD), has been utilized by some manufacturers. Our hypothesis was that since oxygen is not involved in the reaction pathway of glucose dehydrogenase, glucose dehydrogenase-based blood glucose meters would perform better than glucose oxidase-based meters at high altitude. To our knowledge, performance of glucose dehydrogenase-based meters at high altitude and low temperature has not been studied.
Five plasma-calibrated blood glucose meters were evaluated in this study, four glucose dehydrogenase based (GD1: Precision Xtra; GD2: Ascensia Contour; GD3: Accu-Chek Compact; and GD4: Freestyle) and one glucose oxidase based (GO1: OneTouch Ultra), with capillary blood samples from one of the investigators (D.Ö.).
First, all meters were tested in a hypobaric chamber at simulated altitudes (at 20°C in chronological order with ∼8-min intervals) of 0, 4,500, 2,500, and again 0 m above sea level, with normal (∼5.8-mmol/l) and high (∼16.5-mmol/l) plasma glucose values (n = 6 at all conditions). At 4,500 and 2,500 m altitude, the glucose oxidase-based meter (GO1) overestimated plasma glucose values by 15 ± 0.1% (mean ± SD) at the normal blood glucose level and 6.5 ± 0.5% at the high blood glucose level, as compared with the readings at 0 m.
Comparatively, three glucose dehydrogenase–based meters overestimated readings of normal and high blood glucose levels (GD1 by 6.5 ± 0.2 and 1.5 ± 0.7%, GD3 by 3.7 ± 0.1 and 3.5 ± 0.4%, and GD4 by 0.8 ± 0.2 and 0.8 ± 0.4%, respectively). The fourth, GD2, underestimated readings of normal and high blood glucose levels by 1.9 ± 0.2 and 4.2 ± 0.9%, respectively.
Second, the effect of temperature was tested with ∼14.6 mmol/l plasma glucose at high (25°C) and low (8°C) temperature at ground level, allowing 35 min for meters and test strips to acclimate to each temperature. Three meters underestimated, GD1 by 7.9 ± 0.6% (mean ± SD, n = 6), GD2 by 8.5 ± 0.7%, and GD4 by 9.7 ± 0.5%, and two meters overestimated, GO1 by 6.7 ± 0.9% and GD3 by 8.4 ± 0.6%, plasma glucose values at 8°C as compared with 25°C.
In addition, three glucose dehydrogenase–based meters (GD1, GD2, and GD3) were tested with blood at up to 5,895 m above sea level during the ascent of Mount Kilimanjaro, Tanzania. In the presence of both high altitude and low temperature, the meters diverged from each other. At the summit, 5,895 m above sea level, the readings of the investigator’s plasma glucose concentration were 2.8, 11.9, and 21.0 mmol/l (GD1, GD2, and GD3, respectively).
In this study, all four glucose dehydrogenase–based meters performed better than the glucose oxidase–based meter at high altitude, as hypothesized. However, at low temperature, all tested meters performed with similar magnitude of discrepancy. The glucose dehydrogenase–based meters showed a within-group variation, where GD3 alone overestimated plasma glucose levels at low temperature. GD3 determines glucose using reflectance photometry, in contrast to the other (electrochemical) blood glucose meters tested in this study. GD4 performed exceptionally well at simulated high altitude but not at low temperature and is based on coulometric measurement technology, in contrast to the other (amperometric) electrochemical meters (GD1, GD2, and GO1).
In conclusion, people with diabetes who intend to participate in activities at high altitude or, in particular, at low temperature, should be informed that blood glucose meters may give totally unreliable false low or high readings.
Footnotes
- DIABETES CARE