P O S I T I O N S T A T E M E N T

Management of Dyslipidemia in Adults

RATIONALE FOR TREATMENT cholesterol is not significantly increased. tal nor VLDL triglyceride predicted CHD.
OF DYSLIPIDEMIA — The rationale Lastly, as shown in the technical review Baseline data from the United Kingdom
for the treatment of diabetic dyslipidemia (1), the median triglyceride level in type 2 Prospective Diabetes Study (UKPDS)
is discussed in detail in the American Di- diabetic patients is ⬍200 mg/dl (2.30 showed that both decreased HDL and el-
abetes Association (ADA) technical re- mmol/l), and 85–95% of patients have tri- evated LDL predicted CHD (3). In obser-
view “Management of Dyslipidemia in glyceride levels below 400 mg/dl (4.5 vational studies, HDL may be the most
Adults With Diabetes” (1). Type 2 diabe- mmol/l). consistent predictor of CHD in type 2 di-
tes is associated with a two- to fourfold As in nondiabetic individuals, lipid abetes subjects, followed by triglyceride
excess risk of coronary heart disease levels may be affected by factors unrelated and total cholesterol.
(CHD). Although the degree of glycemia to glycemia or insulin resistance, such as
in diabetic patients is strongly related to renal disease, hypothyroidism, and the CLINICAL TRIALS OF LIPID
the risk of microvascular complications frequent occurrence of genetically deter- LOWERING IN DIABETIC
(retinopathy and renal disease), the rela- mined lipoprotein disorders (e.g., familial SUBJECTS — No clinical trial has
tion of glycemia to macrovascular disease combined hyperlipidemia and familial been done on the effects of lipid-lowering
in type 2 diabetes is more modest. The hypertriglyceridemia). These genetic dis- agents on subsequent CHD specifically in
finding of increased cardiovascular risk orders may contribute to the severe hy- diabetic subjects. However, a number of
factors before the onset of type 2 diabetes pertriglyceridemia seen in some patients clinical trials have included small num-
also suggests that aggressive screening for with diabetes. Furthermore, use of alco- bers of adult type 2 diabetic subjects. In
diabetes combined with improved glyce- hol and estrogen may also contribute to the Scandinavian Simvastatin Survival
mic control alone will not be likely to hypertriglyceridemia. Study (4S) trial, simvastatin (HMG CoA
completely eliminate excess risk of CHD reductase inhibitor or “statin”) signifi-
in type 2 diabetic patients. Clearly, a mul- cantly reduced CHD incidence and total
LIPOPROTEIN RISK FACTORS mortality (borderline significantly) in di-
tifactorial approach to prevention of CHD FOR CHD — Relatively few prospec-
in type 2 diabetes will be necessary. abetic subjects with high LDL cholesterol
tive studies of lipids and lipoproteins as and with previous clinical CHD. In the
predictors of CHD have been reported in Cholesterol and Recurrent Events (CARE)
PREVALENCE OF DYSLIPIDEMIA type 2 diabetic subjects, and the results study, pravastatin reduced CHD inci-
IN TYPE 2 DIABETES — The most have been somewhat contradictory. In the dence significantly in diabetic subjects
common pattern of dyslipidemia in type 2 large Multiple Risk Factor Intervention with average LDL levels and with previ-
diabetic patients is elevated triglyceride Trial (MRFIT), total cholesterol as well as ous clinical CHD. In the Helsinki Heart
levels and decreased HDL cholesterol lev- cigarette smoking and blood pressure Study, gemfibrozil (fibric acid derivative)
els. The concentration of LDL cholesterol predicted the development of cardiovas- was associated with a reduction in CHD
in type 2 diabetic patients is usually not cular disease in diabetic and nondiabetic in diabetic subjects without prior CHD
significantly different from nondiabetic subjects, suggesting that risk factors may (although this result was not statistically
individuals. Diabetic patients may have be predictive in both groups. In a Finnish significant). In the Veterans Affairs High-
elevated levels of non-HDL cholesterol study, increased triglyceride levels and Density Lipoprotein Cholesterol Inter-
(LDL plus VLDL). However, type 2 dia- decreased HDL cholesterol levels (but vention Trial (VA-HIT), gemfibrozil was
betic patients typically have a preponder- neither LDL nor non-HDL cholesterol) associated with a 24% decrease in cardio-
ance of smaller, denser LDL particles, predicted CHD in well-characterized type vascular events in diabetic subjects with
which possibly increases atherogenicity 2 diabetes subjects. However, after ad- prior cardiovascular disease (4).
even if the absolute concentration of LDL justment for HDL cholesterol, neither to-
MODIFICATION OF
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● LIPOPROTEINS BY
The recommendations in this paper are based on the evidence reviewed in the following publication: MEDICAL NUTRITION
Management of dyslipidemia in adults with diabetes (Technical Review). Diabetes Care 21:160 –178, 1998. THERAPY AND
The initial draft of this paper was prepared by Steven M. Haffner, MD. This paper was peer-reviewed, PHYSICAL ACTIVITY — The ADA
modified, and approved by the Professional Practice Committee and the Executive Committee, November has made recommendations for both
1997. Most recent review/revision, 2000.
Abbreviations: ADA, American Diabetes Association; CARE, Cholesterol and Recurrent Events Study; medical nutrition therapy (MNT) (5) and
CHD, coronary heart disease: CVD, cerebrovascular disease; MNT, medical nutrition therapy; NCEP, Na- physical activity (6). Weight loss and in-
tional Cholesterol Education Program: 4S, Scandinavian Simvastatin Survival Study. creased physical activity will lead to de-

S74 DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002

Table 1—Category of risk based on lipopro- Table 2—Treatment decisions based on LDL cholesterol level in adults with diabetes
tein levels in adults with diabetes
Medical nutrition therapy Drug therapy
LDL HDL
Initiation level LDL goal Initiation level LDL goal
Risk cholesterol cholesterolⴱ Triglyceride
With CHD, PVD, or CVD ⱖ100 ⬍100 ⱖ100 ⬍100
High ⱖ130 ⬍35 ⱖ400
Without CHD, PVD, and CVD ⱖ100 ⬍100 ⱖ130* ⬍100
Borderline 100–129 35–45 200–399
Data are given in milligrams per deciliter. *For patients with LDL between 100 and 129 mg/dl, a variety of
Low ⬍100 ⬎45 ⬍200 treatment strategies are available, including more aggressive MNT and pharmacological treatment with a
Data are given in milligrams per deciliter. *For statin; in addition, if the HDL is ⬍40 mg/dl, a fibric acid such as fenofibrate may be used in these patients.
women, the HDL cholesterol values should be in- MNT should be attempted before starting pharmacological therapy. PVD, peripheral vascular disease.
creased by 10 mg/dl.
the achievement of optimal glycemic control. used. However, in patients with clinical car-
creased triglyceride and increased HDL Since improved glycemic control may also diovascular disease or very high LDL cho-
cholesterol levels and also to modest low- lower triglyceride levels, it might also cause lesterol levels (i.e., ⱖ200 mg/dl [5.15
ering of LDL cholesterol levels. Diabetic a favorable change in LDL composition. mmol/l]), pharmacological therapy should
patients who are overweight should be be initiated at the same time that behavioral
given a prescription for MNT and for in- TREATMENT GOALS FOR therapy is started.
creased physical activity. The proportion LIPOPROTEIN THERAPY — The In the context of the NCEP report, it is
of saturated fat in the meal plan should be categories of CHD risk by lipoprotein lev- suggested that diabetic subjects with clini-
reduced. The ADA suggests an increase in els in type 2 diabetic patients are shown in cal CHD and an LDL cholesterol level of
either carbohydrate or monounsaturated Table 1. Because of frequent changes in ⱖ100 mg/dl (2.60 mmol/l) after MNT and
fat to compensate for the reduction in sat- glycemic control in diabetic patients and glucose interventions be treated with phar-
urated fat. Some (but not all) studies sug- their effects on levels of lipoprotein, levels macological agents. For diabetic patients
gest that a high–monounsaturated fat diet of LDL, HDL, total cholesterol, and tri- without preexisting CHD, the current ADA
may have better metabolic effects than a glyceride should be measured every year recommendations for starting pharmaco-
high-carbohydrate diet, although other in adult patients. If values fall in lower- logical therapy are 1) an LDL cholesterol
experts have suggested that such a dietary risk levels, assessment may be repeated level of ⱖ130 mg/dl (3.35 mmol/l) and 2) a
modification may make weight loss more every 2 years. In children with diabetes, goal of ⬍100 mg/dl (2.60 mmol/l) for LDL
difficult in obese diabetic patients. consideration should be given to measur- cholesterol. These recommendations are
Recommendations of the American ing lipoproteins after age 2 years, as sug- based not only on the high incidence of
Heart Association for patients with CHD gested by the National Cholesterol CHD in patients with diabetes, but also on
(7) have suggested that the maximal MNT Education Program (NCEP) Report of the their higher case fatality rate once they have
typically reduces LDL cholesterol 15–25 Expert Panel on Blood Cholesterol in CHD. Since a large proportion of diabetic
mg/dl (0.40 – 0.65 mmol/l). Thus, if the Children and Adolescents (8). patients die before they reach the hospital, a
LDL cholesterol exceeds the goal by ⬎25 Optimal LDL cholesterol levels for preventive strategy based solely on second-
mg/dl (0.65 mmol/l), the physician may adults with diabetes are ⬍100 mg/dl (2.60 ary prevention would not be able to “save”
decide to institute pharmacological ther- mmol/l), optimal HDL cholesterol levels are large numbers of these diabetic patients. In
apy at the same time as behavioral therapy ⬎45 mg/dl (1.15 mmol/l), and desirable patients with LDL between 100 mg/dl (2.60
for high-risk patients (i.e., diabetic pa- triglyceride levels are ⬍150 mg/dl (1.7 mmol/l) and 129 mg/dl (3.30 mmol/l), a
tients with a prior myocardial infarction mmol/l). (In women who, at least when variety of treatment strategies are available,
and/or other CHD risk factors). In other nondiabetic, tend to have higher HDL including more aggressive MNT and phar-
patients, behavioral interventions may be cholesterol levels than men, it may be de- macological treatment with a statin (11).
evaluated at 6-week intervals, with con- sirable to have even higher HDL choles- MNT should be attempted before starting
sideration of pharmacological therapy be- terol levels than 45 mg/dl [1.15 mmol/l].) pharmacological therapy. In addition, if the
tween 3 and 6 months. However, raising HDL cholesterol levels HDL is ⬍40 mg/dl, a fibric acid such as
pharmacologically in diabetic patients is fenofibrate might be used in patients with
MODIFICATION OF very difficult since the most effective LDL between 100 and 129 mg/dl.
LIPOPROTEINS BY GLUCOSE- agent raising HDL cholesterol levels is In agreement with the earlier ADA con-
LOWERING AGENTS — Interven- nicotinic acid, which is relatively contra- sensus panel (10), increased triglyceride
tions to improve glycemia usually lower indicated in diabetic patients. Fibrates levels are recognized as a target for interven-
triglyceride levels. In general, glucose-low- can raise HDL cholesterol levels signifi- tion. Since recommended LDL levels are
ering agents do not change or have only a cantly without affecting glycemic control. considered to be ⬍100 mg/dl (2.60 mmol/
modest effect on raising HDL levels. How- The recommendations for treatment of l), and since many diabetic patients have
ever, the HDL composition may change in elevated LDL cholesterol (Table 2) generally increased triglyceride levels, a large propor-
a direction thought to be antiatherogenic. follow the guidelines of both the NCEP (9) tion of diabetic patients will have elevated
Thiazolidinediones may increase HDL and and a recent ADA consensus development levels of both LDL cholesterol and triglycer-
LDL levels, but the long-term effect of such conference (1993) (10), with the following ides. As such, there is likely to be an increase
changes is not known. LDL cholesterol may caveats. Pharmacological therapy should be of diabetic patients on pharmacological
decrease modestly (up to 10 –15%) with initiated after behavioral interventions are therapy and thus an increase in expendi-

DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002 S75

Table 3—Order of priorities for treatment of diabetic dyslipidemia in adults* greater LDL-lowering effects and may be
useful in diabetic patients with combined
I. LDL cholesterol lowering* hyperlipidemia. Although HDL cholesterol,
First choice as noted above, is a powerful predictor of
HMG CoA reductase inhibitor (statin) CHD in diabetic patients, it is difficult to
Second choice raise HDL cholesterol levels without phar-
Bile acid binding resin (resin) or fenofibrate macological intervention. Nicotinic acid,
II. HDL cholesterol raising which should be used with caution in dia-
Behavioral interventions such as weight loss, increased physical activity, and smoking betic patients, and fibrates can effectively
cessation may be useful increase HDL cholesterol levels. Behavioral
Difficult except with nicotinic acid, which should be used with caution, or fibrates interventions (weight loss, smoking cessa-
III. Triglyceride lowering tion, increased physical activity) may in-
Glycemic control first priority crease HDL cholesterol.
Fibric acid derivative (gemfibrozil, fenofibrate) In some cases, combined lipid ther-
Statins are moderately effective at high dose in hypertriglyceridemic subjects who also apy may be initiated. Several options are
have high LDL cholesterol shown in Table 3. The combination of
IV. Combined hyperlipidemia statins with nicotinic acid and especially
First choice with gemfibrozil or fenofibrate has been
Improved glycemic control plus high-dose statin associated with increased risk of myositis,
Second choice although the risk of clinical myositis (as
Improved glycemic control plus statin† plus fibric acid derivative† (gemfibrozil, fenofibrate) opposed to elevated creatinine phos-
Third choice phokinase levels) appears to be low.
Improved glycemic control plus resin plus fibric acid derivative (gemfibrozil, fenofibrate) However, the risk of myositis may be in-
Improved glycemic control plus statin† plus nicotinic acid† (glycemic control must be creased with the combination of gemfi-
monitored carefully) brozil and cerivastatin or in patients with
*Decision for treatment of high LDL before elevated triglyceride is based on clinical trial data indicating safety renal disease. The combination of statins
as well as efficacy of the available agents. †The combination of statins with nicotinic acid and especially with with nicotinic acid is extremely effective
gemfibrozil or fenofibrate may carry an increased risk of myositis. See text for recommendations for patients
with triglyceride levels ⬎400 mg/dl. in modifying diabetic dyslipidemia (with
the largest increases in HDL cholesterol
tures on pharmacological therapy. How- use of combination therapy) is also very ef- levels), but the combination may signifi-
ever, the clinical trial data suggest that fective for reducing triglyceride levels and cantly worsen hyperglycemia. Thus, this
reduction of LDL cholesterol is associated should be aggressively used before the in- combination should be used with ex-
with reduction in CHD and perhaps over- troduction of fibric acids. After the achieve- treme caution: use low doses of nicotinic
all mortality. ment of optimal glycemic control (or at least acid (ⱕ2 g of nicotinic acid per day) with
Economic analyses, based on the 4S after the achievement of as much improve- frequent monitoring of glucose levels.
study, suggest that pharmacological ther- ment as likely to be possible), the physician
apy may be cost-effective once indirect may consider adding a fibric acid. In Table LIPID-LOWERING AGENTS — A
costs of CHD are taken into account (12). 1, the decision to start pharmacological brief summary of the actions of available
Table 3 shows the order of priorities for therapy treatment is dependent on the cli- agents for lipid lowering in patients with
treatment of dyslipidemia. Treatment of nician’s judgment between triglyceride lev- diabetes is shown in Table 4. Generally, one
LDL cholesterol is considered as the first els of 200 mg/dl (2.30 mmol/l) and 400 or two agents are available in each class with
priority for pharmacological therapy of dys- mg/dl (4.50 mmol/l). Above 400 mg/dl the exception of the statins, for which there
lipidemia for a number of reasons (1). Clin- (4.50 mmol/l), strong consideration should are many. The choice of statin should de-
ical trials (4S and CARE) showing the be given to pharmacological treatment of pend principally on the LDL reduction
effectiveness of statins in reducing CHD in triglyceridemia. In contrast, improved gly- needed to achieve the target (⬍100 mg/dl
diabetic subjects show greater risk reduc- cemic control will only modestly reduce [2.60 mmol/l]), on the initial LDL level, and
tions with narrower confidence intervals LDL cholesterol levels, and therefore in di- on the judgment of the treating physician.
than the Helsinki Study with gemfibrozil. abetic patients with both high LDL choles- It should also be noted that the higher
The initial therapy for hypertriglyceri- terol and high glucose levels, one might doses of statins may be moderately effective
demia is behavioral modification with simultaneously initiate glucose lowering at reducing triglyceride levels (though not
weight loss, increased physical activity, and and statin therapy. In some studies, higher- necessarily at raising HDL levels) and thus
moderation of alcohol consumption. In the dose statins are moderately effective in re- may reduce the need for combination ther-
case of severe hypertriglyceridemia ducing triglyceride levels in markedly apy. With the use of high doses of statins,
(ⱖ1,000 mg/dl [11.3 mmol/l]), severe di- hypertriglyceridemic subjects (triglyceride the LDL levels may be reduced to 80 mg/dl
etary fat restriction (⬍10% of calories) (in ⱖ300 mg/dl [3.40 mmol/l]). The critical is- (2.05 mmol/l) or less, and there is no safety
addition to pharmacological therapy) is sue is that gemfibrozil should not be initi- data at such low LDL levels. The use of very
necessary to reduce the risk of pancreatitis. ated alone in diabetic patients who have high-dose statin therapy (i.e., simvastatin
Improved glycemic control (which has undesirable levels of both triglyceride and 80 mg or atorvastatin 40 or 80 mg) to treat
been facilitated by the introduction of new LDL cholesterol. Fenofibrate, a recently ap- hypertriglyceridemia should be restricted to
glucose-lowering agents and more frequent proved fibric acid derivative, may have patients with both high levels of LDL cho-

S76 DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002

Table 4—Pharmacologic agents for treatment of dyslipidemia in adults Manley SE, Matthews DR, Holman RR:
Risk factors for coronary artery disease in
Effect on lipoprotein non-insulin dependent diabetes mellitus
Clinical trials in (UKPDS 23). BMJ 316:823– 828, 1998
LDL HDL Triglyceride diabetic subjects 4. Rubins HB, Robins SJ, Collins D, Fye CL,
First-line agents Anderson JW, Elam MB, Faas FH, Linares
E, Schaefer EJ, Schectman G, Wilt TJ,
LDL lowering
Wittes J: Gemfibrozil for the secondary
HMG CoA reductase inhibitor 22 71 72 4S (simvastatin) prevention of coronary heart disease in
CARE (pravastatin) men with low levels of high-density li-
Triglyceride lowering poprotein cholesterol: Veterans Affairs
Fibric acid derivative 271 1 22 Helsinki (gemfibrozil) High-Density Lipoprotein Cholesterol In-
Second-line agents tervention Trial Study Group. N Engl
LDL lowering J Med 341:410 – 418, 1999
Bile acid binding resins 2 7 1 None 5. American Diabetes Association: Evi-
LDL and triglyceride lowering dence-based nutrition principles and rec-
Nicotinic acid 2 11 22 None ommendations for the treatment and
prevention of diabetes and related com-
In diabetic patients, nicotinic acid should be restricted to ⱕ2 g/day: short-acting nicotinic acid is preferred. plications (Position Statement). Diabetes
Care 25 (Suppl. 1):S50 –S60, 2002
lesterol as well as high triglyceride levels. However, limited data are available 6. American Diabetes Association: Diabetes
Changes in therapy should be done at ⬃4- to from clinical trials, especially in diabetic pa- mellitus and exercise (Position State-
6-week intervals based on laboratory findings. tients without clinical cardiovascular dis- ment). Diabetes Care 24:S51–S55, 2001
ease. In the absence of such data, because of 7. Grundy SM, Balady GJ, Criqui MH, Fletcher
CONSIDERATIONS IN THE the high mortality for diabetic patients with G, Greenland P, Hiratzka LF, Houston-
Miller N, Kris-Etherton P, Krumholz HM,
TREATMENT OF ADULTS first myocardial infarction, aggressive treat- LaRosa J, Ockene IS, Pearson TA, Reed J,
WITH TYPE 1 DIABETES — Type ment of dyslipidemia is also indicated. For Smith SC, Washington R: When to start
1 diabetic patients who are in good control patients without previous CHD, the goal for cholesterol-lowering therapy in patients
tend to have normal (and sometimes better LDL cholesterol is ⱕ100 mg/dl (2.60 with coronary heart disease: a statement for
than normal) levels of lipoprotein. Their mmol/l); the initiation level for pharmaco- healthcare professionals from the American
composition of lipoproteins may be abnor- logical therapy is set at an LDL level ⱖ130 Heart Association task force on risk reduc-
mal, but the effects of these compositional mg/dl (3.35 mmol/l). However, for patients tion. Circulation 95:1683–1685, 1997
abnormalities in relation to CHD are un- with LDL levels between 100 and 129 mg/ 8. Expert Panel on Blood Cholesterol Levels
known. There is relatively little observa- dl, a variety of treatment strategies are avail- in Children and Adolescents: Treatment
tional data on lipoproteins and CHD, and able, including more aggressive MNT and recommendations of the National Choles-
terol Education Program Report of the Ex-
there are no clinical trials relating lipopro- pharmacologicaltreatmentwithastatin.MNT
pert Panel on Blood Cholesterol Levels in
teins to CHD. It seems reasonable that if should be attempted before starting phar- Children and Adolescents. Pediatrics 89
type 1 diabetic patients have LDL choles- macological therapy. In addition, if the HDL (Suppl.):525–584, 1992
terol levels that are above the goals recom- is ⬍40 mg/dl, a fibric acid such as fenofi- 9. Executive Summary of the Third Report
mended for type 2 diabetic patients (Table brate might be used in patients with LDL of the National Cholesterol Education
2), they should be aggressively treated. Im- cholesterol between 100 and 129 mg/dl. Program (NCEP) Expert Panel on Detec-
proved glycemic control may be even more The initial therapy for hypertriglycer- tion, Evaluation and Treatment of High
important in type 1 diabetic patients than in idemia is improved glycemic control. Ad- Blood Cholesterol in Adults (Adult Treat-
type 2 diabetic patients for reduction of ditional triglyceride lowering can be ment Panel III). JAMA 285:2486 –2497,
CHD (e.g., Wisconsin Epidemiologic Study achieved with very high dose statins (for 2001
10. American Diabetes Association: Detection
of Diabetic Retinopathy [WESDR]). subjects with both high LDL and triglyc-
and management of lipid disorders in di-
eride levels) or fibric acid derivatives abetes (Consensus Statement). Diabetes
CONCLUSIONS — Aggressive ther- (gemfibrozil or fenofibrate). Care 16:828 – 834, 1993
apy of diabetic dyslipidemia will probably 11. Haffner SM, Lehto S, Rönnemaa T,
reduce the risk of CHD in patients with Pyörälä K, Laakso M: Mortality from cor-
diabetes. Primary therapy should be di- References onary heart disease in subjects with type 2
rected first at lowering LDL levels. The 1. Haffner SM: Management of dyslipidemia diabetes and in nondiabetic subjects with
goal is to reduce LDL concentrations to in adults with diabetes (Technical Re- and without prior myocardial infarction.
levels recommended for patients with view). Diabetes Care 21:160 –178, 1998 N Engl J Med 339:229 –234, 1998
pre-existing CHD (ⱕ100 mg/dl [2.60 2. Stratton IM, Adler AI, Neil HA, Matthews 12. Herman WH, Alexander CM, Cook JR, Boc-
DR, Manley SE, Cull CA, Hadden D, cuzzi SJ, Musliner TA, Pederson TR, Kjek-
mmol/l]). The initiation level for behav- Turner RC, Holman RR: Association of shus J, Pyorala K: Effect of simvastatin
ioral interventions is also an LDL choles- glycaemia with macrovascular and micro- treatment on cardiovascular resource utili-
terol of ⬎100 mg/dl (2.60 mmol/l)]. The vascular complications of type 2 diabetes zation in impaired fasting glucose and dia-
initial therapy should be to use statin (UKPDS 35): prospective observational betes: findings from the Scandinavian
therapy with the addition of a resin if nec- study. BMJ 321:405– 412, 2000 Simvastatin Survival Study. Diabetes Care
essary to reach the LDL goal. 3. Turner RC, Millns H, Neil HA, Stratton IM, 22:1771–1778, 1999

DIABETES CARE, VOLUME 25, SUPPLEMENT 1, JANUARY 2002 S77