Hyperlipidemia Influences the Accuracy of Glucometer-Measured Blood Glucose Concentrations in Genetically Diverse Mice

Abstract

Background:

Glucometers are widely used in animal research due to simplicity and ease of utilization, but their accuracy in blood glucose assessment for hyperlipidemic mice is unknown.

Methods:

Here, we compared blood glucose levels measured by a glucometer with plasma glucose levels measured by a standard enzymatic assay for 325 genetically diverse F2 mice derived from LP and BALB/c (BALB) Apoe^−/− mice. Non-fasting glucose levels were measured before initiation of a Western diet and after 11 weeks on the diet.

Results:

On chow diet, lab-measured plasma glucose levels were 279.5 ± 42.6 mg/dl (mean ± SD), while blood glucose values measured by glucometer were 138.7 ± 16.6 mg/dl. The two measures had no correlation (R²=0.006, P=0.167). On the Western diet, plasma glucose levels rose to 351.1 ± 121.6 mg/dl, while glucometer-measured blood glucose fell to 128.7 ± 27.9 mg/dl. The two measures showed a moderate correlation (R²=0.111, P=3.1E-9). Lab-measured plasma glucose showed strong positive correlations with plasma triglyceride and non-high-density lipoprotein cholesterol levels, while glucometer-measured blood glucose showed an inverse correlation with non-high-density lipoprotein levels on the chow diet.

Conclusions:

Our results indicate that hyperlipidemia affects the accuracy of glucometers in measuring blood glucose levels of mice.

Introduction

Cardiometabolic syndrome comprises a cluster of metabolic dysfunctions featured by central or abdominal obesity, insulin resistance, dyslipidemia, hyperglycemia, and hypertension, all of which predispose to both cardiovascular diseases and type 2 diabetes (T2D).¹ 80 million Americans, or 1/3 of the adults, are estimated to have metabolic syndrome.¹ The incidence of metabolic syndrome often parallels the incidence of pre-diabetes and T2D. In 2017, 30 million adults, or 12% of American adults, had T2D and 57 million Americans had pre-diabetes, a condition that increases the risk of developing diabetes.² Worldwide, 1 billion adults aged 25 years or older are estimated to have cardiometabolic disorders.³

Fasting hyperglycemia is the criterion currently used to diagnose diabetes and pre-diabetes. A fasting blood sugar level of 126 mg/dl or higher is considered diabetic, and a fasting blood sugar level from 100 to 125 mg/dl is called prediabetes in humans. Blood sugar levels are routinely measured with laboratory biochemical tests to diagnose diabetes and pre-diabetes.² The disadvantages of such tests include the requirement of a relatively large blood volume, the time and expense needed to obtain the results, and the inability to perform the procedure by patients themselves. In contrast, blood glucometers have the advantages of being simple, low cost, and can provide immediate results with use of tiny amounts of blood. Glucometers are also widely used in experimental diabetes research, especially research involving mice.^3,4,5 However, our recent studies have hinted large disparities between the blood glucose levels measured by a glucometer and plasma glucose levels measured by a biochemical assay in apolipoprotein E-deficient (Apoe^−/−) mice.^5,6,7 Apoe^−/− mice are a commonly used animal model for experimental atherosclerosis research. We have observed that Apoe^−/− mice with certain backgrounds develop significant hyperglycemia with fasting plasma levels exceeding 250 mg/dl on a Western diet.^8,5,6 Mice with fasting plasma glucose ≥250 mg/dl are considered diabetic.⁹ This observation has also been replicated by other groups.^10,11

Genetic backgrounds have a dramatic influence on diet-induced hyperglycemia among mouse strains. Apoe^−/− mice with a C57BL/6 or SWR/J genetic background develop significant hyperglycemia but become resistant to it after being transferred onto a BALB/cJ (BALB) background.^6,5 Genetically diverse F2 mice generated from two phenotypically divergent strains in hyperglycemia are expected to exhibit a wide range of variation in blood glucose concentrations that may occur in natural populations. In the present study, we used a F2 cohort to evaluate the accuracy of glucometers in measuring blood glucose levels of hyperlipidemic mice.

METHODS

Mice

BALB-Apoe^−/− and LP-Apoe^−/− mice were generated in our laboratory using the classical congenic breeding approach.¹² Female BALB-Apoe^−/− mice were crossed with male LP-Apoe^−/− mice to generate F1s, which were intercrossed by brother–sister mating to generate 325 F2 mice. Mice were weaned at 3 weeks of age onto a chow diet, switched onto a Western diet containing 21% fat, 0.15% cholesterol, 34.1% sucrose, 19.5% casein, and 15% starch (TD88137, Envigo) at 6 weeks of age, and maintained on the diet for 12 weeks. The animals were housed in a pathogen-free facility with a 12 hour light-12 hour dark cycle. All procedures were carried out in accordance with current National Institutes of Health guidelines and approved by the University Animal Care and Use Committee.

Experimental procedure

Non-fasting sugar levels of F2 mice were measured twice: Once prior to the introduction of the Western diet and once after 11 weeks on the diet. For measurement of blood glucose levels, the tail tip was cut with a clean scissors to release blood, which was applied to a test strip in a glucometer (OneTouch Ultra2 glucometer, LifeScan, Inc.). With a few exceptions, each mouse was checked twice with the glucometer. Immediately following glucometer assessment, blood was drawn from the retro-orbital sinus with mice under isoflurane anesthesia. Ethylenediaminetetraacetic acid (EDTA)-nized blood samples were centrifuged at 13,200 RCF (relative centrifugal force) for 5 minutes, plasma was then collected and stored at −80°C before assay.

Measurements of plasma glucose and lipids

Plasma glucose concentrations were determined with a Sigma assay kit (Cat. # GAHK20), an assay system based on the enzymatic hexokinase oxidase reaction, as previously described⁸ with a modification of 3x dilution for plasma collected from mice fed the Western diet. Briefly, 6 μl of plasma samples, together with standards and controls, were loaded in a 96-well plate and then mixed with 150 μl of assay reagent per well. After a 30-minute incubation at 30 °C, the absorbance at 340 nm was read on a Molecular Devices (Menlo Park, CA, USA) plate reader. Plasma levels of total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglyceride were measured as previously reported,^8,13 except for the cholesterol reagent (Stanbio Laboratory) and high-density lipoprotein (HDL) reagent (FUJIFILM Wako Diagnostics), which produced by different manufacturers. For analysis of total cholesterol levels in mice fed the Western diet, plasma was diluted 3x with H₂O. Non-HDL cholesterol concentrations were calculated as the difference between total and HDL cholesterol levels.

Statistical analysis

Data are expressed as mean ± SD, with n indicating the number of animals. Student’s t test was used to determine the statistical significance for differences between two groups or measurements. Pearson’s correlation analysis was performed to determine correlations between two variables. Differences were considered statistically significant at p ≤ 0.05.

Results

Blood glucose levels measured by glucometer vs. plasma glucose levels measured by lab assay

Blood glucose levels achieved with the glucometer were compared with plasma glucose levels determined by the lab test for the F2 population when fed chow and a Western diet. As shown in Figure 1, the blood glucose value measured by glucometer was only 1/2 the value of plasma glucose measured by the lab test on the chow diet (138.7 ± 16.6 vs. 279.5 ± 42.6 mg/dl; p=3.4E-188). On the Western diet, the value of blood glucose yielded by the glucometer was approximately 1/3 the value of plasma glucose obtained by the lab test (128.7 ± 27.9 vs. 351.1 ± 121.6 mg/dl; p=2.4E-99) (Fig. 1). Compared to the chow diet, the Western diet led to a significant rise in plasma glucose levels of F2 mice (p=8.6E-20) but a significant fall in glucometer-measured blood glucose levels (p=1.1E-07).

Figure 1.

Comparison between blood glucose levels measured by a glucometer and plasma glucose levels measured by an enzymatic assay for F2 mice fed a chow or Western diet. F2 mice were generated from an intercross between BALB- and LP-Apoe^−/− mice. Results are means ± SD (standard deviation) of 300 to 313 mice. Measurements were made immediately before mice were switched onto a Western diet at 6 weeks of age and 11 weeks on the diet. * P < 0.05 vs. plasma glucose levels measured by the enzymatic assay, and ⁺ P < 0.05 vs. measurements when mice were fed the chow diet.

Differences in blood glucose levels between two checkings by glucometer

Blood glucose levels were checked twice by the glucometer for almost all of the F2 mice at either examination (once on chow and once on the Western diet). We compared the blood glucose values between the two checks, and found that the values from the 1^st check were significantly higher than those from the 2^nd one at both occasions: the difference was 20 mg/dl for the chow diet (148.3 ± 18.1 vs 128.6 ± 18.2 mg/dl; p=4.6E-36) and 16 mg/dl for the Western diet (136.6 ± 29.9 vs. 120.4 ± 26.7 mg/dl; p=5.8E-12) (Figure 2).

Figure 2.

Comparison between two readings made by a same glucometer at the same test on blood glucose levels of F2 mice. Only F2 mice with 2 readings at a test were included. Results are means ± SD of 296 to 313 mice. Testing was made once before mice were switched onto a Western diet and once after 11 weeks on the Western diet. * P < 0.05 vs. the 1st reading at the same test by the same glucometer, and ⁺ P < 0.05 vs. measurements made on the chow diet.

We also compared blood glucose levels achieved by the glucometer between the chow and Western diets for the two checks (Figure 2). The blood glucose value on the Western diet was significantly lower than the value on the chow diet for the 1^st (p=7.1E-09) and 2^nd check (p=1.3E-05).

Correlations between blood glucose levels measured by glucometer and plasma glucose levels measured by lab test

There was no significant correlation between the blood glucose values measured by the glucometer and the plasma glucose values measured by the lab test among the F2 mice fed the chow diet (R²=0.006; p=0.167) (Figure 3). On the Western diet, the two measures showed a moderate correlation (R²=0.111; p= 3.2E-09). F2 mice with higher plasma glucose levels tended to have higher blood glucose values on the glucometer.

Figure 3.

Correlations between blood glucose levels measured by a glucometer and plasma glucose levels measured by the enzymatic assay in F2 mice fed a chow or Western diet. Each point represents values of an individual F₂ mouse. The correlation coefficient (R) and significance (p) are shown.

Comparison between glucometer-measured blood glucose and lab-measured plasma glucose levels for associations with plasma lipid variables

On the chow diet, plasma glucose levels measured by the lab test were significantly correlated with plasma levels of triglycerides (R²=0.143; p=2.4E-12) and non-HDL cholesterol (R²=0.150; p=6.3E-13) (Figure 4). F2 mice with higher plasma triglycerides or non-HDL cholesterol levels had higher plasma glucose levels. These correlations were not found with glucometer-measured blood glucose levels; rather, a negative correlation was observed with plasma non-HDL cholesterol levels (R²=0.075; p=9.9E-7). A small but statistical correlation was observed between HDL cholesterol and blood glucose levels (R²=0.039; p=4.1E-4), although there was no correlation between HDL cholesterol and plasma glucose levels (R²=0.001; p=0.507).

Figure 4.

Correlations of plasma glucose levels measured by the enzymatic assay (top row), or blood glucose levels measured by a glucometer (bottom row) with plasma levels of triglyceride, non-HDL (high-density lipoprotein) and HDL cholesterol in F2 mice fed a chow diet. Each point represents values of an individual F2 mouse. The correlation coefficient (R) and significance (p) are shown.

On the Western diet, plasma glucose levels showed strong correlations with plasma levels of triglycerides (R²=0.501; p=1.29E-47) and non-HDL cholesterol (R²=0.353; p=1.67E-30) (Figure 5). In contrast, glucometer-measured blood glucose levels showed a slight correlation with plasma levels of triglycerides (R²=0.0037; p=7.8E-4) but no correlation with non-HDL cholesterol (R²=0.0006; p=0.666). Plasma HDL cholesterol levels showed a slight correlation with both plasma glucose (R²=0.091; p=7.8E-8) and blood glucose levels (R²=0.048; p=1.3E-4), but the magnitude of correlation was larger to the former.

Figure 5.

Correlations of plasma glucose levels measured by the enzymatic assay (top row) or blood glucose levels measured by a glucometer (bottom row) with plasma levels of triglyceride, non-HDL and HDL cholesterol in F2 mice fed the Western diet. Each point represents values of an individual F2 mouse. The correlation coefficient (R) and significance (p) are shown.

DISCUSSION

In this study, we observed large discrepancies of glucometer-measured blood glucose levels from the plasma glucose levels measured by a standard lab test in hyperlipidemic Apoe^−/− mice. The value of blood glucose concentrations measured by the glucometer was only one half the value of plasma glucose concentrations measured by the lab assay on the chow diet and one third of the value of plasma glucose concentrations on the Western diet. The two measures showed no correlation among F2 mice fed the chow diet and a moderate correlation on the Western diet. Even though plasma glucose levels of F2 mice were significantly elevated on the Western diet, glucometer-measured blood glucose levels showed a significant reduction. Moreover, plasma glucose levels were positively correlated with plasma levels of triglycerides and non-HDL cholesterol on both chow and Western diets, while the reverse was observed for the correlation of blood glucose levels measured by the glucometer with plasma non-HDL cholesterol levels on the chow diet.

In current clinical practice, both whole blood and plasma are acceptable for blood glucose assessment. The values of plasma glucose concentrations are approximately 11% higher than the values of blood glucose concentration for same human blood samples.¹⁴ For wild-type male C57BL/6 (B6) mice, blood glucose values measured by glucometers are mostly higher than plasma glucose values measured by a standard lab test for same blood samples obtained from the retroorbital veins or cut tail tips.³ This disparity has also been observed in female B6 mice.⁴ The distribution of glucose in the blood of humans and rodents could be the reason for the discrepant results between humans and mice. Humans have 58% of blood glucose in plasma, while rats have as high as 84%.^15,16 In striking contrast to the previous findings in wild-type mice, the values of glucometer-measured blood glucose concentrations were 50∼63% lower than the values of plasma glucose concentrations in the present study of F2 mice. Hyperlipidemia is an obvious explanation for the discrepant results of Apoe^−/− mice. As F2 mice were generated from Apoe^−/− strains, they were deficient in Apoe and thus developed moderate hyperlipidemia on the chow diet and severe hyperlipidemia on the Western diet. The F2 mice had a plasma cholesterol level of 389 mg/dl on the chow diet and 1195 mg/dl on the Western diet, while wild-type B6 mice have a plasma cholesterol level of <100 mg/dl on a chow diet and ∼200 mg/ml on an atherogenic diet.^17,18 Plasma triglyceride levels of F2 mice on both chow and Western diet were also higher than wild-type B6 mice.^17,18

The evidence supporting the speculation on the interference of hyperlipidemia with the accuracy of glucometers includes: First, plasma cholesterol and triglyceride levels of F2 mice were markedly elevated on the Western diet, while the reverse was observed for the values of blood glucose measured by the glucometer. Second, non-HDL cholesterol levels were inversely correlated with glucometer-measured blood glucose levels among F2 mice fed the chow diet. All glucose assays, including glucometers, use glucose oxidase methods to generate a measurable product proportional to glucose concentrations.¹⁹ The oxidative reaction is vulnerable to interference from redox agents in blood, such as free oxygen species, hemoglobin, and glutathione. Hyperlipidemia and hyperglycemia are major drivers of oxidative stress in Apoe^−/− mice, especially on the Western diet.²⁰ High levels of oxygen species interfere the enzymatic reaction with glucose and causes low glucose readings on glucometers. Lipids in blood may directly affect the reaction with glucose or spectral detection by glucometers.¹⁹

Plasma glucose levels measured by the lab assay were strongly correlated with plasma levels of triglyceride and non-HDL cholesterol among the F2 mice fed either chow or Western diet. The strong correlations between plasma lipids and glucose have also been observed in other genetic crosses derived from Apoe^−/− mouse strains.^12,21 In contrast, the glucometer only revealed a weak correlation between plasma triglyceride and glucose for F2 mice fed the Western diet and a weak negative correlation between non-HDL cholesterol and glucose for the mice fed the chow diet. This result further indicates the inaccuracy of glucometers in assessing blood glucose levels of Apoe^−/− mice.

We observed no correlation between glucometer-measured blood glucose levels and plasma glucose levels measured by the lab test on the chow diet and a week positive correlation between the two measures on the Western diet. This finding provides strong evidence in support of the insufficiency of glucometer assessment for hyperlipidemic mice.

In the present study, we observed that the blood glucose values measured by the glucometer at the second check were significantly lower than the values at the first check even though the difference was moderate (chow: 13.5%; Western: 11.9%). An explanation for the difference between two checks is the expected time-dependent accumulation of hemolysis at the excision site with release of interfering substances from blood cells, such as hemoglobin, NADH, and phosphates that inhibit glucose oxidase activity and affect spectral sensing. Nevertheless, this small difference can make a difference in diagnosing patients with marginal glucose intolerance as diabetes is diagnosed based on cutoff values.

In summary, we used a genetically diverse F2 population to demonstrate the influence of hyperlipidemia on the accuracy of glucometers in measuring blood glucose levels of mice. The blood and plasma glucose values of F2 mice varied widely, covering a wide range of blood and plasma glucose values that could be seen in other sample analyses. Nevertheless, further studies are needed to determine whether the present findings can be extrapolated to all mice strains and/or other animal species.