Diabetes mellitus is a serious and growing health problem worldwide. The disease is caused by a failure of the body to adequately control the concentration of glucose in the blood, which must be kept within narrow limits in order to allow the cells of the body to function properly (Lawal, 2008). In type I diabetes, autoimmune destruction of the insulin-producing pancreatic b-cells leads to a lack of insulin in the blood, the key hormone which regulates blood glucose levels. In contrast, late-onset type II diabetes is characterised by high insulin levels but impaired response of the tissues to this hormone.
Due to the lack of insulin secretion, patients with type I diabetes must measure the glucose concentration of their own blood and inject recombinant exogenous insulin when appropriate (Ellis et al., 2007). It is therefore essential that rapid and accurate methods are available to enable patients to measure their blood glucose concentration. Indeed, portable glucose monitors are now used in a range of different healthcare settings (Montagnana et al., 2009).
These devices use the enzyme glucose oxidase to oxidase glucose, a reaction that can be monitored by a reflectometric technique (colour intensity) or an impedenziometric assay (which measures a change in electrical resistance following glucose oxidation) (Montagnana et al., 2009). However, the performance of these devices is still less accurate than the same method performed in clinical laboratories (Mahoney and Ellison, 2007). This may lead to erroneous or misleading results, which, overall, are thought to negatively impact patient care significantly (Kost et al., 2008). However, the use of this method in the laboratory can be used to produce extremely accurate readings (Montagnana et al., 2009). In this study we evaluated this primary method of measuring blood glucose concentration in the laboratory. The enzyme glucose oxidase specifically catalyses the oxidation of glucose to gluconic acid and hydrogen peroxide. The reaction can be monitored spectrophotometrically at 515 nm, as hydrogen peroxide can be used to oxidise 4-aminophenazone to a magenta-coloured molecule. The specific aims of the experiment were to produce a standard curve to relate the absorbance measured to a range of glucose concentrations, and then to determine the coefficient of variation in the assay and hence the number of replicates required to produce accurate measurements.
Preparation of glucose solutions for the calibration curve: 1 ml glucose was added to 9 ml distilled water to create the first dilution. A serial dilution of this solution was then carried out to produce a range of glucose solutions of different concentrations.
Assay procedure. 0.1 ml of the test solution was mixed with 2.9 ml protein precipitant (this step is necessary to remove the protein present in a blood sample). 3 ml glucose oxidase reagent was then added to initiate the reaction, and the solution was incubated at 37°C for 15 minutes. At the end of the incubation the absorbance of the sample at 515 nm was read using a spectrophotometer.
We first wanted to produce a calibration curve so that the absorbance measured in the assay could be related to the concentration of glucose in the sample. To do this, glucose samples of known concentration were prepared, subjected to the assay procedure and the absorbance measured.
By using this calibration curve, the concentration of glucose in any sample can be obtained by measuring the absorbance of the sample in this assay. The accuracy of the method was then assessed by measuring the glucose concentration of another series of samples, again of known concentration.
Measurement of glucose content of the samples using the assay. The coefficient of variation is defined as the standard deviation divided by the mean, and represents a measure of the variability of the replicates.
In this experiment we have used a simple assay to measure the glucose concentration of a number of samples. The aim of the study was to assess how accurate the method is at measuring glucose concentrations in a sample. Our data (Table 2) indicate that the assay is highly reproducible. The variation between the three replicates we carried out for each sample yielded very similar results; this is illustrated by the low values for the standard deviation and the coefficient of variation. Therefore, three replicates would seem to be sufficient to obtain a very accurate spectrophotometer reading.
In order to assess how accurate the method was at determining the glucose concentration, we can use the calibration curve (Figure 1) to compare the known glucose concentrations of the samples to that determined by our assay (Table 3). This shows that the assay has given extremely inaccurate results, miscalculating the glucose concentration of the samples between approximately 2-fold and 10-fold. However, this is likely due to the poor calibration curve we produced; the readings were taken over too large a range and there were no replicates taken. This meant that the points did not lie on a straight line as we would have expected (the correlation coefficient was low at 0.92), and hence the results of the second part of the experiment were inaccurate. However, judging by the reproducible nature of the data, I would think that the assay could be used to give very accurate readings if the experiment were to be repeated with an improved calibration curve.
The estimated glucose concentrations were determined from the equation of the line of best fit in Figure 1; thus the estimated glucose concentration = (Mean A515 – 0.184) / 20.438.
Therefore, the first target for future work would be to repeat the current experiment and demonstrate that this laboratory assay can give extremely accurate readings when carried out effectively. It may then be interesting to compare the performance of this assay in the laboratory to the assay used by portable glucose monitors, to assess to what extent the performance of these devices is sub-optimal. In the long-term, it is hoped that further technological advances may improve the accuracy of these monitors to eliminate the possibility of dangerous readings (Montagnana et al., 2009).
Appropriately validating biochemical assays used for medical purposes is critical. As diabetics need to keep their blood glucose concentrations within narrow limits, it is crucial that the methods used to measure blood glucose levels are properly tested and validated to make sure they are sufficiently accurate to guide the patient’s insulin injections. This is also true for a whole range of different biochemical assays used in a medical setting.