Several key points emerged from our analyses. First, the lipid profile of adults with undiagnosed diabetes during 1988–1991 was the worst of the four groups; these participants had the lowest adjusted mean concentration of high-density lipoprotein cholesterol or the highest adjusted mean concentrations of total cholesterol, non-high-density lipoprotein cholesterol, triglycerides, and apolipoprotein B. By 2005–2008, however, this group no longer uniformly had the worst mean concentrations of lipids and apolipoprotein B. Second, unadjusted and adjusted mean concentrations of several lipids improved from 1988–1991 to 2005–2008 among adults in all four groups. Third, some of this improvement was likely a consequence of the increased use of cholesterol-lowering medications and, to a lesser degree, the increased use of triglyceride-lowering medications. Fourth, the age-adjusted prevalence of the atherogenic lipid triad, as we defined it, decreased significantly only among adults with prediabetes and normoglycemia.
Adults with diabetes experienced the largest absolute decreases in mean concentrations of total cholesterol, low-density lipoprotein cholesterol, non-high-density lipoprotein cholesterol, and apolipoprotein B. All of these lipids have been shown to be directly related to the risk for cardiovascular disease, although a long-running debate about the superiority of one lipid parameter over another in predicting risk of cardiovascular disease remains unresolved [15–19]. Thus, the improvement in all of these lipid parameters is encouraging and suggests that the risk for cardiovascular disease, at least from lipids, has declined over time. Previous studies have shown that morbidity and mortality from cardiovascular disease among adults with diabetes have declined in the United States [20, 21].
The rapidly increasing use of cholesterol-lowering medications is undoubtedly a major contributor to the improvements in these lipids. The increased use of such medications among adults with diabetes has been reported previously . The contribution of other factors is less clear. The intake of dietary fat in the United States has not changed appreciably in recent decades , and, thus, it seems unlikely that changes in the intake of this macronutrient provide a ready explanation for the observed reduction in mean concentrations of these lipids. One factor that should have worked against favorable improvements in lipid concentrations is the strong increase in body mass index; increased body mass index generally correlates with increased concentrations of low-density lipoprotein cholesterol, non-high-density lipoprotein cholesterol, apolipoprotein B, and triglycerides as well as decreased concentrations of high-density lipoprotein cholesterol. The fact that improvements in concentrations of various lipids were observed in the face of the rising prevalence of obesity testifies to the presence of other powerful secular trends that were able to offset the potentially deleterious impact of obesity on lipid concentrations.
In contrast, the mean concentrations of triglycerides did not change significantly among patients with diabetes. The mean concentration of high-density lipoprotein cholesterol increased significantly only when covariates including body mass index were taken into account. Although the use of medications known to influence concentrations of these lipids increased significantly, especially among adults with diagnosed diabetes, the use of these agents is far less common than that of cholesterol-lowering medications. Hence, the use of fenofibrate, gemfibrozil, and niacin is unlikely to have substantially influenced the observed mean concentrations. Dietary practices can affect concentrations of triglycerides. However, the intake of fat has changed little since 1999–2000 whereas the intake of carbohydrates has decreased . Concentrations of high-density lipoprotein cholesterol are influenced by physical activity, alcohol use, and other factors. Trends in physical activity in the United States remain clouded because of inconsistency in recording physical activity in national surveys. Data from the National Health Interview Survey from 1999 to 2010 shows that some improvement in the percentage of adults meeting current guidelines has taken place . A recent analysis of energy expenditure in the workplace concluded that a reduction in energy expenditure has occurred over five decades . Per capita alcohol consumption has increased slightly from a little over 2.1 gallons during 1997–1998 to about 2.3 gallons during 2009 .
Among people with diabetes, changes in glycemic control can also affect lipid parameters. For example, in a study of 73 patients with type 2 diabetes, improved glycemic control was associated with favorable changes in high-density lipoprotein cholesterol, apolipoprotein A1 and the ratio of apolipoprotein A1 to apolipoprotein B . Because data indicate that glycemic control has improved in the United States [10, 27], this development may have contributed to improving the lipid profiles of people with diabetes.
An interesting observation from our analyses is that adults with undiagnosed diabetes and prediabetes showed evidence of having worse lipid profiles than adults with diagnosed diabetes. The high uptake of cholesterol-lowering medications by adults with diagnosed diabetes likely improved the lipid profile of these adults in comparison with the lipid profiles of adults with undiagnosed diabetes and prediabetes. Nevertheless, mean concentrations of lipids generally changed in a favorable direction among adults with undiagnosed diabetes and prediabetes, although the magnitude of the change was generally smaller than that achieved by adults with diagnosed diabetes. This observation suggests that adults with undiagnosed diabetes and prediabetes are at increased cardiovascular risk from uncontrolled lipid abnormalities and that, once diabetes is diagnosed, adults with previously undiagnosed diabetes will benefit from treatment to manage dyslipidemias.
Apolipoprotein B is the protein that constitutes the principal structural element of lipoprotein particles for very-low-density lipoprotein (VLDL), intermediate-density lipoprotein (IDL), and LDL. Two forms of apolipoprotein B have been identified: apolipoprotein B48 and apolipoprotein B100. The latter is generally measured in assays of apolipoprotein B such as the present study. Apolipoprotein B100 is produced in the liver, incorporated into VLDL particles, and subsequently secreted into the circulation where the particle is transformed into LDL. Because there is one molecule of apolipoprotein B per lipoprotein particle, concentrations of apolipoprotein B reflect the number of VLDL, IDL, and LDL particles. Apolipoprotein B is recognized by cell receptors, leading to the uptake of the lipoprotein particle. Numerous studies have shown that concentrations of apolipoprotein B are directly related to cardiovascular risk, and some have argued that apolipoprotein B is a better lipid measure for assessing cardiovascular risk than are other lipid parameters such as low-density lipoprotein cholesterol and non-high-density lipoprotein cholesterol [15, 17, 18] although considerable disagreement persists [16, 19]. Because apolipoprotein B is not routinely measured in clinical practice at present, alternative methods to derive estimates of concentrations of apolipoprotein B have been pursued .
In persons with diabetes, a limited number of studies has shown that concentrations of apolipoprotein B predict cardiovascular morbidity and mortality [29–32], but not always better than other lipid markers . Concentrations of apolipoprotein B have also been shown to be associated with diabetic retinopathy , microalbumiuria [34, 35], carotid atherosclerosis [36, 37], and coronary artery calcification . Thus, the drop in mean concentration of apolipoprotein B in adults with diabetes in the present study, which was consistent with the decreases in mean concentrations of low-density lipoprotein cholesterol and non-high-density lipoprotein cholesterol, suggests a lessening of cardiovascular risk among people with diabetes. The data suggest that the use of lipid-lowering medications accounted for a substantial portion of the decrease.
Current guidelines from the American Diabetes Association and the American College of Cardiology Foundation call for reducing concentrations of apolipoprotein B to less than 80 mg/dl in people who have diabetes in addition to at least one additional major risk factor for cardiovascular disease . Elevated concentrations of apolipoprotein B were the most pronounced dyslipidemia among all four groups of adults. Pharmacologic approaches to accomplishing this goal in people with diabetes rely mostly on the use of statins . Insulin treatment in patients with type 2 diabetes has been shown to lower concentrations of apolipoprotein B . Potential nonpharmacological approaches includes increased physical activity and dietary change.
The findings of the present study are subject to several limitations. First, methods and laboratories used to conduct the lipid measurements changed during the study period. Although strict quality control procedures were implemented in all surveys, direct comparisons were not done. Supportive of the validity of our observations, however, is that the changes in all the atherogenic indices- low-density lipoprotein cholesterol, non-high-density lipoprotein cholesterol, and apolipoprotein B- were similar. Second, the sample sizes for some of the groups, particularly diagnosed diabetes and undiagnosed diabetes, were small resulting in limited statistical power to detect significant changes in lipid concentrations. The available sample sizes also precluded us from performing stratified analyses by various sociodemographic or other variables.
In conclusion, mean concentrations of several lipids and apolipoprotein B decreased significantly from 1988–1991 to 2005–2008 among adults with diagnosed diabetes. However, no statistically significant changes in concentrations of triglycerides transpired. The rapid increase in the use of lipid-lowering medications, especially among adults with diagnosed diabetes, likely was a factor in the observed changes. Despite these improvements, sizeable percentages of adults with diabetes continue to have elevated concentrations of low-density lipoprotein cholesterol, triglycerides, and especially apolipoprotein B. Thus, further progress in adequately managing dyslipidemias in adults with diabetes, and thereby lessening their risk for cardiovascular morbidity and mortality, remains to be achieved.