Mixed Dyslipidemia Among Patients Using Lipid-Lowering Therapy in French General Practice: An Observational Study

Clinical Therapeutics/Volume 29, Number 8, 2007

Mixed Dyslipidemia Among Patients Using Lipid-Lowering Therapy in French General Practice: An Observational Study Eric Van Ganse, MD, PhD1; Laurent Laforest, PhD1; Thomas Burke, PhD2; Hemant Phatak, PhD2; and Thierry Souchet, MD 3 l Pkarmacoepidemiology Unit, Centre Hospitalier Universitaire-Lyon, Lyon, France; 2Global Outcomes Research, Reimbursement ~ Health TechnologyAssessment, Merck ~ Co., Inc., Wkitehouse Station, NewJersey; and 3Merck Sharp ~ Dohme, Chibret, France ABSTRACT Background: Low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) are associated with an increased incidence of coronary heart disease (CHD). However, limited data are available about the prevalence of dyslipidemias related to LDL-C, HDL-C, and TGs among French patients treated with lipid-lowering agents. Objective: This paper describes the prevalence of various types of dyslipidemias among patients treated with lipid-lowering agents in French general practice. Methods: This was a cross-sectional, observational study conducted using retrospective data collection at the time of enrollment. Eligible patients were those treated pharmacologically for dyslipidemia in the Cegedim Strategic Data general practice network. Fasting lipid values and cardiovascular (CV) risk factors were gathered by investigators using an ad hoc questionnaire. European guidelines were used to define various types of dyslipidemias. Polytomous logistic regression was used to assess the associations between different dyslipidemias and diabetes mellitus, a history of CHD, and the number of CV risk factors. Results: A total of 946 patients had a complete lipid profile and valid data for determining CV risk status. The mean (SD) age of these patients was 64.0 (9.9) years, and 55.7% of the patients were men. At least 1 abnormality in LDL-C, HDL-C, or TGs was present in 791 (83.6 %) of the 946 patients. The rates of elevated LDL-C, low HDL-C, and elevated TGs were 73.2%, 16.9%, and 30.3%, respectively (these groups are not mutually exclusive). Among those who did not reach the LDL-C goal, 38.7% had dyslipidemias with low HDL-C, elevated TGs, or both. Compared with having a normal lipid profile, each additional CV risk factor increased the likelihood of the following types of dyslipidemias: low HDL-C and/or elevated TGs, but August 2007

normal LDL-C (odds ratio [OR], 1.36; 95% CI, 1.03-1.79); elevated LDL-C and TGs, but normal HDL-C (OR, 1.58; 95% CI, 1.24-2.02); and all 3 lipid abnormalities (OR, 1.54; 95% CI, 1.10-2.14). Patients with diabetes had a similarly increased risk of mixed dyslipidemias, whereas patients with a history of CHD did not. Conclusions: Among these patients treated with lipid-lowering agents, 38.7% had mixed dyslipidemias, including low HDL-C, elevated TGs, both low HDL-C and elevated TGs, or all 3 lipid abnormalities. Patients with a greater number of nonlipid CV risk factors or with diabetes had a significantly increased risk of mixed dyslipidemias involving elevated TGs and/or low HDL-C in addition to elevated LDL-C. (Clin Ther. 2007;29:1671-1681) Copyright © 2007 Excerpta Medica, Inc. Key words: dyslipidemia, low-density lipoprotein cholesterol, LDL-C, high-density lipoprotein cholesterol, HDL-C, triglycerides, TGs, primary care.

INTRODUCTION

Low-density lipoprotein cholesterol (LDL-C) is a wellestablished risk factor for coronary heart disease (CHD). The beneficial impact of therapeutic LDL-C lowering with statins on CHD risk reduction has been established, 1-4 and statins are the first-line pharmacologic therapy for lowering LDL-C. 5 High-density

Portions of this research were presented at the 2007 European Atherosclerosis Meeting, June 10-13, 2007, Helsinki, Finland.

Acceptedfor publicationJune26, 200Z doi:l 0.1016/].clinthera.2007.08.003 0149-2918/$32.00 Printed in the USA. Reproduction in whole or part is not permitted.

Copyright © 2007 Excerpta Medica, Inc.

1671

Clinical Therapeutics

lipoprotein cholesterol (HDL-C) is inversely associated with cardiovascular (CV) risk. 6-s Some trials have found an independent association between HDL-C increases from fibrate or statin therapy and CHD risk reduction. 1,9,1° At present, elevation of HDL-C is not a treatment goal in European or French guidelines, ua2 although thresholds for low HDL-C have been identified in European guidelines. 12 The evidence for an independent association between triglycerides (TGs) and CHD risk is mixed. Some studies have reported an effect of elevated TGs on CHD risk, independent of other lipid abnormalities, 13-2° although this association was weaker than the associations of LDL-C and HDL-C with CHD. 21,22 Limited data are available on the prevalence of dyslipidemias related to LDL-C, HDL-C, and TGs in French patients who are using lipid-lowering agents (LLAs). Identifying dyslipidemias prevalent in specific population subgroups, such as those with diabetes, CHD, and multiple CV risk factors, may aid in the prevention of cardiovascular disease (CVD) and improve the quality of care. For example, a significant association between mixed dyslipidemias and high CV risk may prompt consideration of therapies targeted at mixed dyslipidemias. 12 Identification of parameters associated with mixed dyslipidemia may assist in the management of these patients through pharmacologic or nonpharmacologic means. Therefore, the objective of this study was to describe the prevalence of different types of lipid disorders, particularly low HDL-C and hypertriglyceridemia, in patients treated with LLA therapy. We also evaluated the association between patients' characteristics, including diabetes, history of CHD, multiple CV risk factors, and different types of dyslipidemias. PATIENTS A N D METHODS Design and Data Source A cross-sectional study was performed using data from the Cegedim Strategic Data (CSD) network, a French database of general practitioners (GPs). This database allows linking of patients' clinical data with their responses to an ad hoc questionnaire administered by the participating GPs using a computer program at the time of an inclusion visit. 23 The questionnaire was used to capture laboratory data and information about the presence of various CHD risk factors. The participating GPs activated the computer screens when patients agreed to participate. An 1672

anonymous identification number allowed linking of patient-level data collected on the computer screens to the information obtained from CSD. In the original study,24 348 (20.5%) GPs were randomly selected from 1698 GPs in the database with the aim of enrolling 3000 patients. Of the 348 randomly selected GPs, 136 participated in the study and 123 (35.3%) ultimately enrolled 3173 patients. Patients were administered the ad hoc questionnaire only once, on the day of enrollment; there was no follow-up. The study protocol was approved in writing by the Comit8 National Informatiques et LibertG the independent French body that oversees all observational studies performed locally. Patient Selection Details of patients' inclusion criteria have been described elsewhere.24 Briefly, patients were included in the original study if they were diagnosed with a dyslipidemia and were supervised by GPs belonging to the CSD network. Patients were selected if a statin or fibrate prescription was recorded for that patient in the CSD database. Specifically, the study criteria required that patients had been prescribed an LLA (fibrate or statin) between July 2001 and June 2002. The index prescription was the first LLA prescription during this period. In addition, a previous prescription for an LLA (not necessarily the same as the index prescription) was required to be recorded in the CSD database at least 6 months before the index prescription. The previous prescription could have occurred before July 2001. The study enrollment period lasted from September 2002 to October 2002. Patients were required to have the results of at least 1 fasting blood lipid panel within 12 months before the inclusion visit. As per the protocol, when several tests were available for a patient during the 12 months before the inclusion visit, only the most recent results were used to generate a lipid profile for that patient. In summary, GPs enrolled patients into this crosssectional study after confirming eligibility, which included a diagnosis of dyslipidemia, treatment with an LLA (current and previous prescriptions), and availability of blood lipid laboratory results. Data presented in this study focus on patients from the original study24 for whom LDL-C, HDL-C, and TG values were available and for whom participating physicians collected a complete medical history as required to evaluate patients' CV risk status based on Volume 29 Number 8

E. Van Ganse

guidelines from the Agence Frangaise de S&urit8 Sanitaire des Produits de Sant8 (AFSSAPS; hereafter referred to as the French guidelines), u The French guidelines require 2 occasions with an elevated lipid profile to support a diagnosis of dyslipidemia. Because our patients were experienced with LLAs, we used only 1 profile, the most recent results before enrollment. This decision was made a priori.

Definition of Dyslipidemia European guidelines were used to identify patients with high LDL-C, low HDL-C, and elevated TGs. 12 Elevated LDL-C was defined as LDL-C >2.5 mmol/L in patients with diabetes and/or clinically established CVD, and LDL-C >3 mmol/L in other patients. Low HDL-C was defined as <1.0 mmol/L for men and <1.2 mmol/L for women. Hypertriglyceridemia was defined as a fasting TG level >1.7 mmol/L. 12 A comparison between the European guidelines and the US guidelines 8 (National Cholesterol Education Program [NCEP]) is provided in Table I.

et

al.

Definitions CHD and CV risk factors were defined according to the French guidelines, u These risk factors include age (men >50 years and women >60 years), family history of early severe CHD, smoking status, hypertension, diabetes, and low HDL-C (<1 mmol/L). Four primary-prevention categories were defined based on the French guidelines: no CV risk factors and 1, 2, or _>3 risk factors. An additional category was defined as any patient with a history of CVD, complicated type 2 diabetes, or a 10-year CV risk of >20%. Patients with a history of CVD had stable or unstable angina, myocardial infarction (MI), silent documented MI, and/or vascular diseases including stroke and stage II peripheral arteriopathy. CHD was defined as any of the foregoing, excluding stroke and vascular diseases. Complicated diabetes included any patient with type 2 diabetes and either kidney problems (proteinuria >300 mg/24 hours; creatinine clearance [CockcroftGault formula] <60 mL/min) or at least 2 of the following risk factors: age _>50 years for men or _>60years

Table I. Comparison between European and US guidelines for cardiovascular risk prevention. Lipid Parameter

European Guidelines 12

US Guidelines: NCEP ATP III 8~

Elevated LDL-C

>2.5 mmol/L in patients with diabetes and/or clinically established CVD; >3 mmol/L in remaining patients

_>100 mg/dL (_>2.5 mmol/L) near or above optimal (initiate drug therapy if lifestyle changes are not adequate) _>130 mg/dL (_>3.1 mmol/L)in patients with CHD or CHD risk equivalent, including diabetes _>130 mg/dL (_>3.1 mmol/L) for patients with _>2 risk factors and 10-year CHD risk of 10% to 20% _>160 mg/dL (_>4.0 mmol/L) in patients with _>2 risk factors and 10-year CHD risk <10% _>190 mg/dL (4.9 mmol/L) in patients with 0 to 1 risk factor

Low HDL-C

<1.0 mmol/L for men and <1.2 mmol/L for women

<40 mg/dL (<1.0 mmol/L)

Elevated TGs

FastingTGs >1.7 mmol/L

Borderline-high TGs: 150-199 mg/dL (1.72.3 retool/L) High TGs: 200-499 mg/dL (2.3-5.6 mmol/L) Very high TGs: _>500 mg/dL (_>5.6 mmol/L)

NCEP ATP III = National Cholesterol Education Program Adult Treatment Panel III; LDL-C = Iow-densie/lipoprotein cholesterol; CVD = cardiovascular disease; CHD = coronary heart disease; HDL-C = high-densie/lipoprotein cholesterol; TGs = triglycerides. *Major risk factors in the NCEP ATP III guidelines include cigarette smoking, hypertension (_>140/90 mm Hg or taking antihypertensive medications), low H DL-C, family history of premature CH D (CH D in male first-degree relatives <55 years; CH D in female first-degree relatives <65 years), and age (men _>45years and women _>55years).

August 2007

1673

Clinical Therapeutics

for women, history of early CHD (MI or sudden death before the age of 55 years for the father or a first-degree male relative or before the age of 65 years for the mother or a first-degree female relative), current smokers or past smokers who quit smoking in the past 3 years, treated or untreated hypertension, HDL-C <1.0 retool/L, or microalbuminuria >30 mg/d. Patients with high CV risk had a projected 10-year CV risk of >20% as assessed by the modified Framingham CV risk equation. 25

Statistical Analysis Descriptive analyses evaluated baseline patient characteristics and the prevalence of elevated LDL-C, elevated TGs, and low HDL-C (not mutually exclusive). The prevalence of dyslipidemia was also evaluated in the subgroups with CVD, diabetes, and a projected 10-year primary CVD risk of >20%. Further descriptive analyses evaluated the prevalence of 8 mutually exclusive lipid disorders, defined as all possible combinations of normal or abnormal LDL-C, HDL-C, and TGs, with the use of thresholds described in the 2003 European guidelines. 12 The 8 categories included isolated elevated LDL-C, isolated low HDL-C, isolated elevated TGs, elevated LDL-C and TGs, elevated LDL-C and low HDL-C, low HDL-C and elevated TGs, all 3 lipid abnormalities, and no lipid abnormality. Polytomous logistic regression models were used to evaluate the association between diabetes, CHD history, and the number of CV risk factors with the likelihood that a patient would experience a mutually exclusive lipid disorder versus no lipid disorder. 26,27 Covariates included in these models were age, smoking status, hypertension, family history of an early CV event, diabetes, CHD, and statin monotherapy. Most patients with dyslipidemia in France are treated with statins or fibrates, as represented in this study. The choice of treatment was taken into consideration when calculating the adjusted odds of experiencing a particular type of mixed dyslipidemia for each highrisk group (diabetes, CHD, or high CV risk). SAS version 9.0 (SAS Institute Inc., Cary, North Carolina) was used for all analyses.

RES U LTS Study Population A complete lipid profile and valid data for determining CV risk status according to the AFSSAPS clas1674

sification were available for 946 (29.8%) of the 3173 patients enrolled in the original study (Table II). The mean (SD) age was 64.0 (9.9) years, and 55.7% of the patients were men. Hypertension and diabetes were present in 70.2% (664) and 23.6% (223) of the study population, respectively. More than half of the study population (51.2% [484]) was at high CV risk according to the revised AFSSAPS classification. Approximately one third (33.5% [317]) of patients had a history of CVD. Statin monotherapy was used by 67.1% (635) of patients and fibrate monotherapy by 31.8% (301). Mean (SD) values for LDL-C, HDL-C, and TGs were 3.30 (0.97) retool/L, 1.46 (0.46) retool/L, and 1.59 (1.16) retool/L, respectively.

Prevalence of Dyslipidemia In this cross-sectional study conducted using retrospective data collection at the time of enrollment, 73.2% (692/946) of the study population was not at the therapeutic goal for LDL-C (Table HI). Among patients with elevated LDL-C, 38.6% (267/692) also had low HDL-C and/or elevated TGs. Of these 267 patients, 39.3% (105) had a history of CVD and 30.3% (81) had diabetes. One hundred forty-eight patients with elevated LDL-C had not had a CV event and thus qualified for primary CV prevention. Of these, 45 (30.4%) patients had a 10-year CV risk of >20%. For LDL-C, 26.8% (254/946) of patients were at or below their goal. Of these patients, 39.0% (99/254) had low HDL-C and/or elevated TGs, and among these 99 patients, 41.4% (n = 41) had a history of CVD and 32.3% (n = 32) had diabetes. Fifty-five patients qualified for primary CV prevention, of whom 8 (14.5%) had a 10-year CV risk of >20%. Overall, regardless of LDL-C status, the prevalence of low HDL-C was 16.9% and the prevalence of elevated TGs was 30.3%. Prevalence of Mutually Exclusive Lipid Disorders Most (83.6% [791/946]) of the patients treated with LLAs had at least 1 lipid disorder. The prevalence of mutually exclusive lipid disorders in LLA-treated patients is summarized in Table IV. Among the total study population of 946 patients, 499 (52.7%) had at least 1 of the 3 isolated lipid disorders, and 292 (30.9%) had multiple lipid disorders. Overall, 38.7% (366) of the study population experienced either isolated low HDL-C, isolated elevated TGs, or any mixed dyslipidemia. Finally, 5.9% (56) of the population had all 3 lipid disorders, and 99 (10.5%) had normal LDL-C values but Volume 29 Number 8

E. Van Ganse et al.

Table II. Demographic and clinical characteristics of the French study population treated for dyslipidemia.

Study Population (N = 946)

Original Study Population 24 (N = 31 73)

Age, mean (SD), y

64.0 (9.9)

65.8 (10.8)

Men, no. (%)

527 (55.7)

1609 (50.7)

Weight, mean (SD), kg*

75.7 (14.7)

74.3 (14.9)

Variable

CV risk factors, no. (%) Current smoker Diabetes mellitus Hypertension Family history of early CH D

125 223 664 178

(13.2) (23.6) (70.2) (18.8)

373 529 1965 513

(11.8) (16.7) (61.9) (16.2)

High CV risk (by AFSSAPS criteria), no. (%) History of CVDt High-risk diabetest Any of the above criteria History of CH D

31 7 194 484 257

(33.5) (20.5) (51.2) (27.2)

835 444 1168 666

(26.3) (14.0) (36.8) (21.0)

Current lipid-lowering therapy, no. (%) Statins alone Fibrates alone

Combination therapy (statin + other agent)§ Blood lipid levels, mean (SD), mmol/L LDL-C HDL-C TGs

635 (67.1) 301 (31.8) 10 (1.1)

1892 (59.6) 1217 (38.4) 64 (2.0)

3.30 (0.97) 1.46 (0.46)

3.31 (0.94) 1.50 (0.50)

1.59 (1.16)

1.57 (1.27)

cv = cardiovascular; CH D = coronary heart disease; AFSSAPS = Agence Frangaise de S@curit@Sanitaire des Produits de Sant4; CVD = cardiovascular disease; LDL-C= low-density lipoprotein cholesterol; HDL-C= highdensity lipoprotein cholesterol; TGs = triglycerides. *Weight was available for 2566 of the initial sample and for 879 of the 946 patients included in the analyses (range, 43-137 kg). fCVD includes CHD and ischemic stroke. tHigh-risk diabetes was defined using the AFSSAPS criteria of either concomitant presence of at least 2 CV risk factors or renal disease. §Combination therapy does not indicate systematic concomitant use.

elevated TGs and/or low HDL-C. It is of interest that isolated LDL-C elevation continued to be a problem in 53.7% (425/791) of the patients with lipid-profile disorders despite the use of lipid-lowering agents. Association Between Dyslipidemia and Diabetes, Coronary Heart Disease, and Number of Cardiovascular Risk Factors The results suggested an association between diabetes and dyslipidemia and between multiple CV risk August 2007

factors and dyslipidemia (Table V). Among the 3 lipids studied in patients with diabetes, the strongest association was found for elevated TGs in combination with either elevated LDL-C (odds ratio [OR], 1.49; 95% CI, 1.11-1.97) or low HDL-C (OR, 1.44; 95% CI, 1.06-1.96). The presence of multiple CV risk factors (including diabetes, age, family history of early severe CVD, hypertension, and smoking) was significantly associated with elevated TGs in combination with either elevated LDL-C (OR, 1.58; 95% CI, 1675

Clinical Therapeutics

Table III. Prevalence ofdyslipidemia ~ based on goal attainment for low-density lipoprotein cholesterol (LDL-C) in the total French study population and subgroups.

% of LLT-Treated Population Dyslipidemia Categoryf

(N = 946)

Patients not at LDL-C goal: 73.2% (n = 692) Among patients not at LDL-C goal, 38.6% (n = 267) also had low HDL-C and/or elevated TGs 39.3% (n = 105) had a history of CVD 30.3% (n = 81) had diabetes 30.4% (n = 45) had a 10-year CV risk >20%t Among patients not at LDL-C goal, 61.4% (n = 425) had only LDL-C abnormality and normal HDL-C and TGs Patients at LDL-C goal: 26.8% (n = 254) Among patients at LDL-C goal, 39.0% (n = 99) had low HDL-C and/or elevated TGs 41.4% (n = 41) had a history of CVD 32.3% (n = 32) had diabetes 14.5% (n = 8) had a 10-year CV risk >20%§ Among patients at LDL-C goal, 61.0% (n = 155) had no other dyslipidemia: normal TGs and normal HDL-C

73.2 28.2 11.1 8.6 4.8

44.9 26.8 10.5 4.3 3.4

0.8 16.4

LLT = lipid-lowering therapy; HDL-C = high-density lipoprotein cholesterol; TGs = triglycerides; CVD = cardiovascular disease; CV = cardiovascular. *Categories are not mutually exclusive. (Thresholds according to the European guidelines are as follows: high TGs, >1.7 mmol/L; low HDL-C, <1.0 mmol/L for men and <1.2 mmol/L for women; LDL-Cnot at goal, >2.5 mmol/L for patients with diabetes and/or clinically established CVD and >3 mmol/L for patients without diabetes and without clinically established CVD. tThere were 148 primary-prevention patients with elevated LDL-Cand low HDL-C and/or elevated TGs. §There were 55 primary-prevention patients with normal LDL-Cand low HDL-C and/or elevated TGs.

1.24-2.02), low H D L - C (OR, 1.36; 95% CI, 1.03-1.79), or both (OR, 1.54; 95% CI, 1.10-2.14). DISCUSSION The French primary care patients enrolled in this study had a surprisingly high frequency of lipid abnormalities, despite receiving LLAs: 73.2% had elevated LDL-C, 16.9% had low HDL-C, and 30.3% had elevated TGs. Attainment of therapeutic goals for LDL-C and of optimal levels of HDL-C and TGs was poor. u J2 Elevated TGs and low HDL-C were present in 38.7% of the population, regardless of LDL-C levels. Most (83.6%) of the patients had at least 1 lipid abnormality despite using LLAs. Among these, 53.7% did not achieve LDL-C goals as defined by current guidelines. 11,12 Compared with our previous results, 24 the present study found a higher rate of nonattainment of the therapeutic goal for LDL-C. One possible 1676

reason may be that physicians preferentially ordered complete lipid profiles for patients who were not at LDL-C goal. In a population-based study in 3281 patients aged 20 to 79 years, almost 70% of the patients did not achieve the NCEP-specified LDL-C goals. 28 However, our study results cannot be compared directly with those obtained using NCEP thresholds because of the differences in threshold values and CV risk categories, as well as possible differences in treatment algorithms between US and French physicians. The mean HDL-C level in our study sample was similar to levels reported in other similar studies. 29-31 The median HDL-C level of 1.4 mmol/L in a sample of untreated French patients 29 is very similar to the mean (SD) HDL-C value of 1.46 (0.46) mmol/L in our study. A similar mean HDL-C level (1.42 retool/L) was reported in the 1998 Health Survey for England. 3° Our study found a slightly greater prevalence of low HDL-C Volume 29 Number 8

E. Van Ganse et al.

Table IV. Prevalence of different lipid disorders in the French study population treated with lipid-lowering agents (N = 946). Lipid Disorder ~

No. (%)

No lipid disorder

1,55 (16.4)

At least 1 lipid disorder

791 (83.6)

Isolated lipid disorder Elevated LDL-C Low HDL-C Elevated TGs

42s (44.9) 21 (2.2) s3 (s6)

Any of the 3 isolated lipid disorders

499 ($2.7)

Mixed lipid disorders Elevated LDL-C, low HDL-C, normal TGs Elevated LDL-C, elevated TGs, normal HDL-C Elevated TGs, low HDL-C, normal LDL-C Elevated LDL-C, elevated TGs, low HDL-C

lS3 (16.2)

Any of the 4 mixed lipid disorders

292 (30.9)

s8 (6.1)

2s (2.6) s6 (s9)

LDL-C= low-densitylipoprotein cholesterol; HDL-C= highdensity lipoprotein cholesterol;TGs = triglycerides;CVD = cardiovascular disease. *Dyslipidemiathresholds are based on the European guidelines: high TGs, >1.7 mmol/L; low HDL-C, <1.0 mmol/L for men and <1.2 mmol/Lfor women; LDL-Cnot at goal, >2.5 mmol/L for patients with diabetes and/or clinically established CVD and >3 mmol/Lfor patients without diabetes and without clinically established CVD.

among treated patients (16.9%) than did Ferrieres et a129 in untreated French patients (12.0%). A reduced incidence of additional CHD events has been observed after improvement in HDL-C levels with LLAs in CHD patients. 32 The importance of raising HDL-C is evident from the recent recommendation by a European Consensus Panel to achieve an HDL-C level of at least 1.03 mmol/L in patients with CHD, patients with a CHD-equivalent condition, or those at high risk of a CHD event. 33 Optimal HDL-C targets recommended in the literature for diabetic men and women are >1.02 and >1.2 mmol/L, respectively. 34 In this study, 30.3% of patients had elevated TGs. Moreover, among 791 patients with at least I lipid abnormality, 287 (36.3%) had elevated TGs. Mounting August 2007

evidence from a number of sources supports an independent association between hypertriglyceridemia and CHD. For example, a meta-analysis based on 17 studies identified TGs as a CV risk factor for both men and women. 17 The Copenhagen Male Study 16 also showed a significant association between TGs and ischemic heart disease (P < 0.05), independent of other risk factors including HDL-C. A joint effect of low HDL-C and elevated TGs is also important in association with CV events. 35 Our results indicated that mixed dyslipidemias occurred more frequently in diabetic patients and in those with a greater number of CV risk factors. Regardless of the comorbid conditions studied, the need for comprehensive lipid management (LDL-C, HDL-C, and TGs) stems from the fact that diabetic patients and patients with a greater number of risk factors are more likely to have mixed dyslipidemia as a residual lipid abnormality than isolated elevation in LDL-C. Compared with nondiabetic patients, those with diabetes in our study had a greater risk of experiencing dyslipidemia involving TGs and HDL-C (Table V). This finding is supported by observations of increased TGs and low HDL-C in other studies involving diabetic patients. 3<37 In a European study of diabetic patients from Austria, Belgium, Finland, France, Germany, Italy, The Netherlands, Portugal, Sweden, Switzerland, and the United Kingdom, those treated with LLAs had a higher prevalence of low HDL-C and elevated TGs. 3s Overall, our findings highlighted the need to target HDL-C and TGs among patients with diabetes. 39 The study results also indicated that a higher number of CV risk factors was associated with an increased likelihood of mixed dyslipidemia as opposed to a normal lipid profile. Our findings are similar to those of a Canadian study, which found that the prevalence of elevated TGs (>2.3 mmol/L) and low HDL-C (<0.9 mmol/L) increased with an increasing number of CV risk factors, including smoking, hypertension, body mass index >27 kg/m2, sedentary lifestyle, and LDL-C >3.4 mmol/L.36 As in other studies, 24,39 clinical targets for lipid management related to LDL-C, HDL-C, or TGs were achieved by few patients in our study, particularly by few of the patients at high CV risk. Further investigations should assess plausible causes for mixed dyslipidemia, including the limited availability of effective therapeutic options to treat low HDL-C and/or elevated TGs or the limited impact of guidelines 1677

Clinical

Therapeutics

Table V. Likelihood of a lipid disorderversus no lipid disorder according to the presence of diabetes, coronary heart disease (CHD), and number of cardiovascular risk factors in the French study population (N = 926). U nadjustedt Subgroup and Lipid Disorder*

OR

Diabetes mellitus§ No lipid disorder Isolated elevated LDL-C Elevated TGs and/or low HDL-C, normal LDL-C Elevated LDL-C, low HDL-C, normal TGs Elevated LDL-C, elevated TGs, normal HDL-C Elevated LDL-C, elevated TGs, low HDL-C

1.0 1.11 1.,54 1.32 1.,51 1.,53

History of CHDII No lipid disorder Isolated elevated LDL-C Elevated TGs and/or low HDL-C, normal LDL-C Elevated LDL-C, low HDL-C, normal TGs Elevated LDL-C, elevated TGs, normal HDL-C Elevated LDL-C, elevated TGs, low HDL-C No. of cardiovascular risk factorsll No lipid disorder Isolated elevated LDL-C Elevated TGs and/or low HDL-C, normal LDL-C Elevated LDL-C, low HDL-C, normal TGs Elevated LDL-C, elevated TGs, normal HDL-C Elevated LDL-C, elevated TGs, low HDL-C

1.0

Adjustedt

9,5% CI

0.87-1.41

1.14-2.07 0.92-1.89 1.1 ,5-1.98 1.08-2.18 -

0.96

0.77-1.19

1.24 1.18 1.28 1.16

0.94-1.64 0.84-1.64 1.00-1.64 0.83-1.63

1.0 1.0,5 1.40 1.29 1.63 1.,54

0.87-1.27

1.08-1.83 0.94-1.76 1.29-2.07 1.11-2.12

OR

9,5% CI

1.0

-

1.09 1.44 1.22 1.49 1.43

0.8,5-1.40 1.06-1.96 0.84-1.76 1.11-1.97 0.99-2.06

1.0

-

0.94 1.11 1.16 1.08 1.04

0.7,5-1.18 0.83-1.49 0.81-1.64 0.83-1.40 0.73-1.48

1.0

-

1.07 1.36 1.26 1. , 5 8 1. , 5 4

0.88-1.30 1.03-1.79 0.91-1.76 1.24-2.02 1.10-2.14

OR = odds ratio; LDL-C= low-density lipoprotein cholesterol; TGs = triglycerides; HDL-C = high-density lipoprotein cholesterol; CVD = cardiovascular disease. *Dyslipidemia thresholds are based on the European guidelines: high TGs, >1.7 mmol/L; low HDL-C, <1.0 mmol/L for men and <1.2 mmol/L for women; LDL-C not at goal, >2.5 mmol/L for patients with diabetes and/or clinically established CVD and >3 mmol/L for patients without diabetes and without clinically established CVD. t Z2 analysis without adjustment for covariates. * Polytomous logistic regression. §Covariates included CHD, diabetes, smoking status, hypertension, age, family history of early severe CHD, and statin monotherapy. IICovariates included CHD, statin monotherapy, and number (count) of cardiovascular risk factors (including age, family history of early severe CVD, smoking status, hypertension, and diabetes).

that lack targets for TGs and HDL-C. In this study, 98.9% of the patients were treated using either statin or fibrate monotherapy, which indicates a heavy emphasis on LDL-C management and is consistent with the current prescribing behavior of physicians in France. 4° Considering the high prevalence of mixed dyslipidemia in these treated patients and the recognition of low H D L - C as a major risk factor according to the European and NCEP guidelines, the use of niacin1678

based therapies in addition to existing treatments may help in lipid management. N o significant association was found between a history of C H D and an increased risk of any type of dyslipidemia. One explanation may be that patients with C H D were managed more aggressively than patients with C H D risk-equivalent conditions. It is also possible that C H D patients were more adherent to both pharmacologic and nonpharmacologic treatVolume

29

Number

8

E. Van Ganse et al.

ments. Finally, patients with more severe CHD may have sought the services of specialists, leading to an overrepresentation of patients with less severe CHD in our GP study. This study had some limitations. Dosages of LLAs were not taken into account, and patients' actual adherence to LLAs was unknown. We focused on patients treated with 1 LLA; only 10 patients received 2 LLAs. At the time this study was conducted, most of the statins were branded and LLA monotherapy was the typical treatment in France. Fibrates are also commonly used in France. Furthermore, the possible effects of diet and exercise on lipid management were not studied. Because complete lipid profiles and valid data for risk classification were not available for all patients, only 29.8% (946/3173) of patients from our original sample were included in this study. This fact may also result in selection bias, because patients with high CV risk tend to have better documentation and hence a greater probability of selection in the cohort. Baseline characteristics of these patients were similar to those in the complete cohort of 3173 patients treated with LLAs, except for a greater prevalence of diabetes and hypertension among the 946 participants. 24 The greater percentage of diabetic patients may also explain the greater rate of TG testing or complete lipid-profile testing in these patients by GPs participating in this study. The requirement of blood pressure values for computation of the Framingham risk score may have resulted in a greater prevalence of hypertension among patients enrolled in this study. Our subsample of 946 patients was similar to but not strictly representative of the remaining 2227 original study 24 patients who did not meet all inclusion criteria for the present study. Despite some differences in CV risks, the numbers of CV risk factors (age >50 years for men or age >60 years for women, hypertension, smoking, diabetes, and family history of CVD) were comparable: 2.1 for the 946 patients in this study versus 1.8 for the 2227 patients from the original study who were excluded from this study. Nonetheless, one should use caution when extrapolating these data to the general population of patients using LLAs. CONCLUSIONS

Our study found low HDL-C levels and elevated TGs in addition to elevated LDL-C levels among these LLA-treated French primary care patients, particularAugust 2007

ly among those at high risk for a CV event. Nearly 4 in 10 (38.7%) patients treated with LLAs had low HDL-C and/or elevated TGs. Patients with diabetes or a high number of CV risk factors had a significantly increased risk of mixed dyslipidemia. ACKNOWLEDGMENTS

Analyses for this study were supported by Merck & Co., Inc. Thomas Burke and Hemant Phatak are employees of Merck & Co., Inc. REFERENCES 1. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: The Scandinavian Simvastatin Survival Study (45). Lancet. 1994;344:1383-1389. 2. Shepherd J, Cobbe SM, Ford I, et al, for the West of Scotland Coronary Prevention Study Group. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N EnglJ Med. 1995;333:1301-1307. 3. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: A randomized placebo-controlled trial. Lancet. 2002;360:7-22. 4. Baigent C, Keech A, Kearney PM, et al, for the Cholesterol Treatment Trialists' (CTi-) Collaborators. Efficacy and safety of cholesterol-lowering treatment: Prospective meta-analysis of data From 90,056 participants in 14 randomised trials ofstatins [published correction appears in Lancet. 2005;366:1358]. Lancet. 2005;366:1267-1278. 5. Stone NJ, BilekS, Rosenbaum S. Recent National Cholesterol Education Program Adult Treatment Panel III update: Adjustments and options. Am J Cardiol. 2005;96: 53E-59E. 6. Gordon DJ, ProbstfieldJL, Garrison RJ, et al. High-density lipoprotein cholesterol and cardiovascular disease. Four prospective American studies. Circulation. 1989;79:8-15. 7. Grundy SM, Cleeman Jl, Merz CN, et al, for the Coordinating Committee of the National Cholesterol Education Program. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III Guidelines. J Am Coil Cardiol. 2004; 44:720-732. 8. National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106: 3143-3421. 9. RobinsSJ. Targeting low high-densitylipoprotein cholesterol for therapy: Lessonsfrom the Veterans Affairs High-density

1679

Clinical Therapeutics

Lipoprotein Intervention Trial. A m J

Cardiol. 2001 ;88:19N-23 N. 10. Wilhelmsen L, Pyorala K, Wedel H, et al. Risk factors for a major coronary event after myocardial infarction in the Scandinavian Simvastatin Survival Study (4S). Impact of predicted risk on the benefit of cholesterollowering treatment. Eur Heart J. 2001 ;22:1119-1127. 11. Agence Francaise de S&urit~ Sanitaire des Produits de Sant& Therapeutic management of dyslipidemic patients: Guidelines [in French[. Saint Denis, France: French Health Products Safety Agency (AFSSAPS); 2005. Available at: http://www. afssaps.sante.fr/pdf/ 5/rbp/ dysreco. pdf. Accessed November 20, 2006. 12. De Backer G, Ambrosioni E, BorchJohnsen K, et al. European guidelines on cardiovascular disease prevention in clinical practice: Third joint task force of European and other societies on cardiovascular disease prevention in clinical practice (constituted by representatives of eight societies and by invited experts). Eur J Cardiovasc Prey Rehabil. 2003;10:$1-$10. 13. Criqui MH, Heiss G, Cohn R, et al. Plasma triglyceride level and mortality from coronary heart disease. N EnglJ Med. 1993;328:1220-1225. 14. Avins AL, Neuhaus JM. Do triglycerides provide meaningful information about heart disease risk? Arch Intern Med. 2000; 160:1937-1944. 15. Menotti A, Scanga M, Morisi G. Serum triglycerides in the prediction of coronary artery disease (an Italian experience). Am J Cardiol. 1994;73:29-32. 16. Jeppesen J, Hein HO, Suadicani P, Gyntelberg F. Triglyceride concentration and ischemic heart disease: An eight-year follow-up in the Copenhagen Male Study [published correction appears in Circulation. 1998; 97:1995]. Circulation. 1998;97:10291036. 17. Hokanson JE, Austin MA. Plasma triglyceride level is a risk factor for car-

1680

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

diovascular disease independent of high-density lipoprotein cholesterol level: A meta-analysis of populationbased prospective studies.J Cardiovasc Risk. 1996;3:213-219. Bainton D, Miller NE, Bolton CH, et al. Plasma triglyceride and high density lipoprotein cholesterol as predictors of ischaemic heart disease in British men. The Caerphilly and Speedwell Collaborative Heart Disease Studies. BrHeartJ. 1992;68:60-66. Stampfer MJ, Krauss RM, MaJ, et al. A prospective study of triglyceride level, low-density lipoprotein particle diameter, and risk of myocardial infarction.JAMA. 1996;276:882-888. Iso H, Naito Y, Sato S, et al. Serum triglycerides and risk of coronary heart disease among Japanese men and women. AmJ Epidemiol. 2001; 153:490-499. Psaty BM, Anderson M, Kronmal RA, et al. The association between lipid levels and the risks of incident myocardial infarction, stroke, and total mortality: The Cardiovascular Health Study.JAm Geriatr Soc. 2004; 52:1639-1647. Haim M, Benderly M, Brunner D, et al. Elevated serum triglyceride levels and long-term mortality in patients with coronary heart disease: The Bezafibrate Infarction Prevention (BIP) Registry. Circulation. 1999;100: 475-482. Van Ganse E. Use of computerized data in pharmacoepidemiology [in French[. Therapie. 2000;55:123-126. Van Ganse E, Souchet T, Laforest L, et al. Ineffectiveness of lipid-lowering therapy in primary care. BrJ Clin Pharmacol. 2005;59:456-463. Wilson PW, D'Agostino RB, Levy D, et al. Prediction of coronary heart disease using risk factor categories. Circulation. 1998;97:1837-1847. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 1989. Kleinbaum DG, Klein M. Logistic Regression: A SelfLearning Text. 2nd ed. New York, NY: Springer; 2002.

28. Persell SD, Lloyd-Jones DM, Baker DW. Implications of changing national cholesterol education program goals for the treatment and control of hypercholesterolemia. J Gen Intern Med. 2006;21:171-176. 29. FerrieresJ, RuidavetsJB, Perret B, et al. Prevalence ofdyslipidaemias in a representative sample of the French population [in French]. Arch Mal Coeur Vaiss. 2005;98:127-132. 30. Primatesta P, Poulter NR. Lipid concentrations and the use of lipid lowering drugs: Evidence from a national cross sectional survey. BMJ. 2000;321 : 1322-1325. 31. Carroll MD, Lacher DA, Sorlie PD, et al. Trends in serum lipids and lipoproteins of adults, 1960-2002.JAMA. 2005;294:1773-1781. 32. Robins SJ, Collins D, WittesJT, et al, for the VA-H IT Study Group (Veterans Affairs High-Density Lipoprotein Intervention Trial). Relation of gemfibrozil treatment and lipid levels with major coronary events: VA-HIT: A randomized controlled trial.JAMA. 2001 ; 285:1585-1591. 33. Chapman MJ, Assmann G, Fruchart JC, et al, for the European Consensus Panel on HDL-C. Raising high-density lipoprotein cholesterol with reduction of cardiovascular risk: The role of nicotinic a c i d - a position paper developed by the European Consensus Panel on HDL-C. CurrMedRes Opin. 2004;20:1253-1268. 34. Pogach LM, Brietzke SA, Cowan CLJr, et al. Development of evidence-based clinical practice guidelines for diabetes: The Department of Veterans Affairs/Department of Defense guidelines initiative. Diabetes Care. 2004;27 (Suppl 2):B82-B89. 35. Jeppesen J, Hein HO, Suadicani P, Gyntelberg F. Relation of high TG-Iow HDL cholesterol and LDL cholesterol to the incidence ofischemic heart disease. An 8-year follow-up in the Copenhagen Male Study. Arterioscler Thromb VascBiol. 1997;17:1114-1120. 36. Connelly PW, Petrasovits A, Stachenko S, et al, for the Canadian

Volume 29 Number 8

E. Van Ganse e t al

37.

38.

39.

40.

Heart Health Surveys ResearchGroup. Prevalence of high plasma triglyceride combined with low HDL-C levels and its association with smoking, hypertension, obesity, diabetes, sedentariness and LDL-C levels in the Canadian population. CanJCardiol. 1999; 15:428-433. Verges B. New insight into the pathophysiology of lipid abnormalities in type 2 diabetes. Diabetes Metab. 2005; 31:429-439. Bruckert E, Baccara-Dinet M, McCoy F, Chapman J. High prevalence of low HDUcholesterol in a panEuropean survey of 8545 dyslipidaemic patients. Curr Med Res Opin. 2005;21:1927-1934. Hobbs R, Hoes A. Effective management ofdyslipidaemia among patients with cardiovascular risk: Updated recommendations on identification and follow-up. EurJ Gen Pract. 2005;11:6875. Van Ganse E, Laforest L, Alemao E, et al. Lipid-modifying therapy and attainment of cholesterol goals in Europe: The Return on Expenditure Achieved for Lipid Therapy (REALITY) study. Curr Med Res Opin. 2005;21 : 1389-1399.

Address c o r r e s p o n d e n c e to: Eric Van Ganse, M D , PhD, Unit6 de Pharmaco6pid6miologie, H6pital Pierre Wertheimer, 59 bd Pinel, F - 6 9 6 7 7 Bron Cedex, France. E-marl: [email protected] August 2007

1681