sacImgMarkCss">Figure 5. Data from regression analysis showing change in glucose vs change in LDL (n = 31) with an R value of 0.484.
Figure 6. Data from regression analysis showing change in HgbA1c vs change in LDL (n = 67) with an R value of 0.527. *Hemoglobin A1c; †Low density lipoprotein.
Figure 7. Data from regression analysis showing change in glucose vs change in HDL (n = 33) with an R value of –0.567. †High density lipoprotein.
Figure 8. Data from regression analysis showing change in Hg-bA1c vs change in HDL (n = 77) with an R value of –0.135. *Hemoglobin A1c; †High density lipoprotein.
Table 2. Predicted changes in all lipid modalities in the glucose intolerant patient group with an increase in glucose by 1 mg/dL in the glucose intolerant group and an increase in hemoglobin A1c by 1 percent in the diabetic group.
be made between these two patient groups due to differrences in the scales of the independent variable (glucose vs. hemoglobin A1c).
Our data shows that there is a statistically significant relationship between disorders of insulin resistance and dyslipidemias, therefore supporting the existing guidelines to screen patients with glucose intolerance for the various dyslipidemias. Both patient populations in our study show statistically significant relationships between changes in glucose or hemoglobin A1c and changes in the various components of the lipid panel over the same time period. The only component that did not show statistical significance in both groups was that of HDL, yet there was a somewhat strong inverse relationship between those two variables in the glucose intolerant patient group.
Our research shows that the strongest relationship was seen with comparisons involving triglycerides in both patient groups. Both groups have R values in excess of 0.70. This strong relationship supports the basic science literature, which states that one of the first components of a lipid panel to increase is that of the triglyceride level secondary to lipoprotein lipase dysfunction. This increase in triglycerides helps to create the subsequent increases of total cholesterol, LDL, and decrease in HDL. Our study does show that increases in LDL and total cholesterol do occur, yet not to the extent of triglycerides. This can be expected considering that the triglyceride level is the most affected component and that elevations in triglyceride levels help create the increases in LDL and total cholesterol.
In regard to the total cholesterol and LDL components, our results show a strong relationship in both patient groups. This helps to reinforce the concept stating that insulin resistance does predispose a patient to hyperlipidemia.
HDL is the only component in which our data does not show strong relationships in both patient groups. Analyses of the patients in the glucose intolerant group show a somewhat strong inverse relationship between HDL and changing glucose levels. This helps to reinforce the idea that decreasing HDL is associated with insulin resistance. This, ultimately, aids in the progression of atherosclerotic coronary artery disease. Yet, the analyses involving patients in the diabetic population show a very weak relationship in regard to changing HDL levels. This relationship does not support the claim that increased levels of insulin resistance help maintain a decreased amount of HDL in the blood. According to the literature, HDL levels are affected most downstream in the aforementioned pathways [1-3]. Therefore, it can be expected that the changes in HDL will not be as strong as the other lipid components that are affected earlier. HDL levels are also one of the more difficult to control and follow over time, especially since many other external factors can manipulate a patient’s HDL level. Further research would be recommended as follow up to investigate this finding in the diabetic patient population.
Both groups have very similar R values when looking at the relationships between change in glucose or hemoglobin A1c and triglyceride, total cholesterol, and LDL levels. Such comparable results between both patient groups help to show the significance of our data. It also suggests that the concept of diabetic dyslipidemia can be applied to patients with varied levels of insulin resistance, therefore reinforcing the need to screen insulin resistant nondiabetic patients.
One interesting finding is that the patients in the glucose intolerant group had higher average levels of total cholesterol, LDL, and triglycerides when compared to the diabetic patient group. To the same extent, this group also has lower levels of HDL. It would have been predicted that such patients have lower overall values in total cholesterol, LDL, and triglycerides and higher levels of HDL, since the diabetic patients have increased levels of insulin resistance. One explanation to these findings is that patients in the diabetic group are likely to be on more stringent medication regimens. Such patients are deemed to be at a higher baseline risk for cardiovascular disease and are treated as such. As well, there are fewer patients and data points in the glucose intolerant group, which could add to a potential observational bias.
One of the study’s limitations is the limited amount of data points that were included. To avoid biases in data, our strict exclusion criteria reduced the number of data points substantially. Patients with both a diagnosis of either diabetes or glucose intolerance and hyperlipidemia with accurate medication records were only used. To the same extent, any changes in medication or gaps in data excluded patients. Even though charts of all active patients in a busy lipid clinic were reviewed, not all charts included complete records that were required for inclusion in our study, which ultimately limited our sample size.
Another limitation is the lack of information regarding lifestyle changes and other behavioral modalities that can affect both glucose intolerant disorders and hyperlipidemia. The lipid institute has information regarding appointments with its dietician, yet information about behavior and lifestyle modifications that a patient may have otherwise undertaken outside of the lipid institute was not used to determine eligibility in the study.
Our study helps to show that diabetic dyslipidemia is a true concern in regard to atherosclerotic coronary artery disease. Increasing glucose and hemoglobin A1c levels do, in fact, create an atherogenic environment. In patients with metabolic syndrome, this further compounds the problem and increases the severity. Other studies have linked diabetes to atherosclerosis by looking at patients already diagnosed with heart disease, yet our study makes the direct link between insulin resistance and dyslipidemia. It is one of the few studies that gives statistical evidence that diabetes and hyperglycemic states directly affect lipid levels.
As discussed, the ATP III Guidelines point out that diabetes is a modifiable factor when treating patients with diabetic dyslipidemia. It also states that the risk of coronary heart disease is substantially higher in diabetic patients when compared with non-diabetic patients and remarks that it is appropriate to place patients with diabetes in a separate risk category than those without . Our study reinforces this concept by showing the strong relationships between those patients with insulin intolerance and the various dyslipidemias.
To that same extent, the American Heart Association considers diabetes to be one of the major controllable risk factors for cardiovascular disease . Various clinical studies, including the Heart Protection Study and Collaborative Atorvastatin Diabetes Study, have shown that there is a true reduction of cardiovascular events in diabetic patients treated with statins for dyslipidemia [7,8].
Our results help to reinforce the need to be vigilant in the treatment of hyperlipidemia in patients that are hyperglycemic. Based on the results, it can be suggested that lipid disorders be routinely tested for in both diabetic and non-diabetic insulin resistant patients and that treatment should be strongly considered if lipid levels are not at goal. It should be noted that insulin resistant states in general predispose patients to dyslipidemias, yet a patient classified to be diabetic is at a much higher risk of both metabolic syndrome and dyslipidemia than a patient with glucose intolerance that has not met the classification of a true diabetic state.
To the same effect, diet and exercise should continuously be reinforced in hyperglycemic patients to help prevent the onset of such lipid disorders, as well as prevent metabolic syndrome from further complicating the problem. Patients with a family history of diabetes should also be strongly encouraged to start participating in diet and behavioral modifications in an attempt to prevent the onset of insulin resistance, which will ultimately lead to dyslipidemias. Our study also suggests that coronary artery disease is a major concern in diabetic and glucose intolerant patients, more so than the general population.