Summary
Background
Heterozygous familial hypercholesterolaemia (HeFH) is a severe autosomal dominant disease that is underdiagnosed, inadequately treated and has a severe long-term cardiovascular risk. Few studies have evaluated the long-term risk of high low-density lipoprotein cholesterol (LDL-C) concentrations.
Aim
To evaluate long-term mortality in a large cohort of healthy subjects, according to LDL-C concentrations.
Methods
Based on a sample of 6956 subjects visiting a preventive cardiology department, we selected adult subjects without a personal history of cardiovascular disease. From 1995 to 2011, 4930 healthy subjects were examined and followed up until 31 December 2011. All-cause deaths were collected exhaustively. A Cox-based multivariable analysis evaluated long-term total mortality risk according to Dutch Lipid Clinic Network (DLCN) LDL-C concentrations.
Results
After a mean follow-up of 8.6 years, 123 all-cause deaths were recorded (cumulative mortality rate, 2.5%). In the final multivariable model, major risk factors such as age, sex, tobacco use and diabetes were significantly associated with mortality. After adjustment for age, sex, tobacco use, hypertension, diabetes and statin therapy, and in comparison with subjects with LDL-C < 4 mmol/L (< 155 mg/dL), subjects with LDL-C between 4 and < 5 mmol/L (155 to < 190 mg/dL) had a hazard ratio (HR) of 1.99 (95% confidence interval [CI] 1.31–3.02; P = 0.001), subjects with LDL-C between 5 and < 6.5 mmol/L (190 to < 250 mg/dL) had an HR of 1.81 (95% CI, 1.06–3.02; P = 0.030), subjects with LDL-C between 6.5 and < 8.5 mmol/L (250 to < 330 mg/dL) had an HR of 2.69 (95% CI, 1.06–6.88; P = 0.038) and subjects with LDL-C ≥ 8.5 mmol/L (≥ 330 mg/dL) had an HR of 6.27 (95% CI, 0.84–46.57; P = 0.073). After excluding patients on statins at baseline, subjects with LDL-C ≥ 8.5 mmol/L (≥ 330 mg/dL) had an HR of 8.17 (95% CI, 1.08–62.73; P = 0.042).
Conclusions
The severity of LDL-C elevation is associated with a higher risk of death in healthy subjects. DLCN LDL-C concentrations may be used in daily practice to identify patients with HeFH who warrant aggressive treatment.
Résumé
Contexte
L’hypercholestérolémie familiale hétérozygote (HeFH) est une maladie autosomique dominante sévère. L’HeFH est une maladie qui est sous-diagnostiquée et insuffisamment traitée. Le risque cardiovasculaire à long terme de l’HeFH est sévère. Peu d’études ont évalué le risque à long terme des valeurs élevées de LDL-cholestérol (LDL-C).
Objectif
Le but de ce travail a été d’évaluer la mortalité à long terme d’une large cohorte de sujets sains en fonction des niveaux de LDL-C.
Méthodes
À partir d’un échantillon de 6956 patients examinés dans un service de cardiologie préventive, nous avons isolé les sujets majeurs sans antécédents personnels de maladie cardiovasculaire. De 1995 à 2011, 4930 sujets sains ont été examinés et suivis jusqu’au 31 décembre 2011. L’ensemble des décès toutes causes a été collecté. Une analyse multivariée par méthode de Cox a permis d’évaluer le risque de mortalité totale à long terme en fonction des niveaux de LDL-C du Dutch Lipid Clinic Network (DLCN).
Résultats
Après un suivi moyen de 8,6 ans, 123 décès toutes causes ont été enregistrés (mortalité cumulative de 2,5 %). Dans le modèle multivarié final, les facteurs de risque majeurs tels que l’âge, le sexe, le tabagisme et le diabète sont associés significativement à la mortalité. Après ajustement pour l’âge, le sexe, le tabac, l’hypertension artérielle, le diabète et un traitement par statines et en comparaison avec les sujets dont le LDL-C est < 4 mmol/L (< 155 mg/dL), les sujets présentant un LDL-C entre 4 et < 5 mmol/L (155 à < 190 mg/dL) ont un hazard ratio (HR) de 1,99 (intervalle de confiance [IC] à 95 %, 1,31–3,02 ; p = 0,001], les sujets présentant un LDL-C entre 5 et < 6,5 mmol/L (190 à < 250 mg/dL) ont un HR de 1,81 (IC à 95 %, 1,06–3,02 ; p = 0,030), les sujets présentant un LDL-C entre 6,5 et < 8,5 mmol/L (250 à < 330 mg/dL) ont un HR de 2,69 (IC à 95 %, 1,06–6,88 ; p = 0,038) et les sujets dont le LDL-C est ≥ 8,5 mmol/L (≥ 330 mg/dL) ont un HR à 6,27 (IC à 95 %, 0,84–46,57 ; p = 0,073). Après exclusion des patients sous statines à l’entrée, les sujets dont le LDL-C est ≥ 8,5 mmol/L (≥ 330 mg/dL) ont un HR à 8,17 (IC à 95 %, 1,08–62,73 ; p = 0,042).
Conclusions
La sévérité de l’élévation du LDL-C est liée à une augmentation de la mortalité totale chez les sujets sains. Les niveaux de LDL-C de la DLCN peuvent être utilisés en pratique quotidienne pour sélectionner les patients porteurs d’une HeFH et ainsi engager une thérapeutique agressive.
Background
Heterozygous familial hypercholesterolaemia (HeFH) is an autosomal dominant disorder with a severe cardiovascular prognosis. A definite diagnosis of HeFH requires genetic testing for mutations primarily located on the low-density lipoprotein (LDL) receptor, apolipoprotein B and proprotein convertase subtilisin/kexin type 9 (PCSK9). Patients with HeFH are perceived as being at high risk, according to the European Society of Cardiology . The European Society of Atherosclerosis recommends extensive screening for HeFH in the general population . Screening of a given subject allows the whole family to be assessed and leads to management of all those exposed to hypercholesterolaemia .
HeFH is a severe disease because of its well-documented cardiovascular and survival prognosis . As a definite diagnosis is based on testing for mutations , this genetic diagnostic test is appropriate for the majority of subjects with elevated LDL-cholesterol (LDL-C) concentrations. Nevertheless, most developed countries are not equipped with sufficient genetic centres to conduct extensive diagnostic tests in subjects for whom this diagnosis has been suggested. Hence, scientific societies mostly propose the use of scores, such as the USA MEDPED (Make Early Diagnosis to Prevent Early Death) score, the UK Simon Broome score and the Dutch Lipid Clinic Network (DLCN) score . With the exception of the USA score, the other scores are based on a collection of clinical and laboratory data, which are not necessarily found in the medical records.
The DLCN score is based on a combination of five dimensions. The first group corresponds to a first-degree relative with known premature (< 55 years, men; < 60 years, women) coronary heart disease (CHD) (1 point), or a first-degree relative with known LDL-C > 95th percentile by age and sex for country (1 point), or a first-degree relative with tendon xanthoma and/or corneal arcus (2 points), or a child/children aged < 18 years with LDL-C > 95th percentile by age and sex for country (2 points). The second group corresponds to premature (< 55 years, men; < 60 years, women) CHD (2 points) or premature (< 55 years, men; < 60 years, women) cerebral or peripheral vascular disease (1 point). The third group corresponds to the presence of tendon xanthoma (6 points) or corneal arcus in a person aged < 45 years (4 points). The fourth group corresponds to LDL-C concentrations: ≥ 8.5 mmol/L (≥ 330 mg/dL) (8 points); ≥ 6.5 to < 8.5 mmol/L (≥ 250 to < 330 mg/dL) (5 points); ≥ 5.0 to < 6.5 mmol/L (≥ 190 to < 250 mg/dL) (3 points); ≥ 4.0 to < 5.0 mmol/L (≥ 155 to < 190 mg/dL) (1 point). The fifth group corresponds to deoxyribonucleic acid (DNA) analysis and the finding of a causative mutation (8 points). A “definite HeFH” diagnosis can be made if the subject scores > 8 points. A “probable HeFH” diagnosis can be made if the subject scores 6 to 8 points. A “possible HeFH” diagnosis can be made if the subject scores 3 to 5 points. An “unlikely HeFH” diagnosis can be made if the subject scores 0 to 2 points. However, this long list of variables is not usually found in current clinical charts. So, we sought to propose a simpler tool for screening for HeFH in general practice.
The aim of this research was to determine whether DLCN LDL-C concentrations affect the survival prognosis in healthy subjects, and may thus be used in the general population.
Methods
Study population
We conducted a prospective cohort study, which included 6956 apparently healthy asymptomatic subjects. Participants were included between November 1995 and December 2011 at the Department of Preventive Cardiology in our teaching institution (Toulouse University Hospital, Toulouse, France). These subjects were either self-referred or referred by their primary-care physician or their cardiologist for cardiovascular risk assessment, management of cardiovascular risk factors or routine ambulatory screening for cardiovascular diseases (CVDs) . Patients with a personal history of coronary heart disease (International Classification of Disease, 9th revision, codes 410.0 to 414.9), a personal history of stroke (codes 433.0 to 438.9 except 437.3 to 437.7), a personal history of atherosclerosis (codes 440.0 to 440.9) or a personal history of aneurysms (codes 441.0 to 442.9) were excluded. Vital status on 31 December 2011 was obtained for each participant through the national database that records all deaths occurring in the French population each year (RNIPP) . Authorisation to use these data was obtained in accordance with French law (Commission nationale de l’informatique et des libertés [CNIL]).
Questionnaires and measurement of clinical variables
At baseline, extensive questionnaires were filled in by trained medical staff during a personal interview with the participant. Information on exposures was collected at baseline only. Data concerning socioeconomic level, personal medical history, cardiovascular risk factors, lifestyle habits and drug intake were recorded. Participants were asked to bring their latest drug prescription to the inclusion visit. All drugs taken during the 2 weeks preceding the visit were recorded. Family history of premature CVD (before 55 years in father/65 years in mother) was recorded. People who currently smoked or who had stopped for < 3 years were considered as current smokers. Height, weight and arterial blood pressure (mean of two measurements performed with a standard sphygmomanometer in a seated position after ≥ 5 minutes rest) were measured according to standardised protocols by the medical staff. Body mass index was calculated as weight divided by height squared (kg/m2). Hypertension was assessed for people with blood pressure ≥ 140/90 mmHg or on treatment. Diabetes was assessed for subjects receiving hypoglycaemic drugs or with fasting blood glucose ≥ 7 mmol/L.
Laboratory methods
Blood samples were taken after ≥ 10 hours of overnight fasting. Serum total cholesterol and triglycerides were measured by enzymatic assays (Boehringer, Mannheim, Germany). High-density lipoprotein cholesterol (HDL-C) was measured after sodium phosphotungstate-magnesium chloride precipitation of apolipoprotein B-containing lipoproteins. LDL-C was determined by the Friedewald formula when triglycerides were < 4.6 mmol/L (< 400 mg/dL) . Glucose concentrations were measured using a conventional enzymatic method based on hexokinase-glucose-6-phosphate dehydrogenase.
Statistical analysis
Statistical analysis was performed using STATA statistical software, release 11.2 (STATA Corporation, College Station, TX, USA). Subjects with a history of CVD were excluded from the analysis.
We first described the baseline characteristics of participants and compared baseline characteristics by outcome occurrence, comparing subjects who did not die (i.e. those alive on 31 December 2011) with subjects who had a fatal event during follow-up. Qualitative variables were compared between groups using the χ 2 test (or Fisher’s exact test when necessary). Student’s t -test was used to compare the distribution of quantitative data (or the Mann–Whitney test when distribution departed from normality or when homoscedasticity was rejected).
Survival was then analysed. Events were cases of death, and exposure was defined by LDL-C concentration at inclusion. Hazard ratios (HRs) for mortality and 95% confidence intervals (CIs) were assessed using a Cox model. The independent variables initially introduced into the survival model were LDL-C concentrations at inclusion and all variables associated with mortality in the univariate analysis (P-value < 0.20). A backward analysis was then applied until only variables significantly and independently associated with mortality (P-value < 0.05) remained. Hypertension was included in the final model because it is a classical risk factor. The proportional-hazard assumption was tested for each covariate by the “log-log” plot method curves (–ln{–ln[survival]}), for each category of nominal covariate, versus (ln[analysis time]). None of the assumptions could be rejected.
Results
A total of 6956 subjects visited the Department of Preventive Cardiology from November 1995 to December 2011. After excluding minors and patients with CVD at inclusion, together with patients who returned to the same department on several occasions, we reached a total of 4930 subjects who could be followed up until 31 December 2011 ( Table 1 ).
| Number of patients | |
|---|---|
| Total examined up to 31 December 2011 | 6956 |
| Examined for the first time | 5182 |
| Minors (< 18 years) | 10 |
| History of ischaemic heart disease | 131 |
| History of stroke | 66 |
| History of documented atherosclerosis | 34 |
| History of vascular aneurysm | 3 |
| Undocumented blood pressure | 4 |
| Lipid profile not performed | 4 |
| Followed up | 4930 |
| Cumulative number of deaths among those followed up | 123 |
After a mean follow-up period of 8.6 years, 123 deaths were recorded (cumulative mortality rate of 2.5%). Among the 4930 healthy subjects at baseline, LDL-C could not be calculated in 106 patients (including two deaths) because of excessively high triglyceride concentrations.
The clinical and laboratory characteristics of the study population are shown in Table 2 . The mean age of subjects was 52 years, 59.4% were men, 25.3% were regular smokers and 5.2% were diabetic. At inclusion, 26.8% of patients were receiving statin therapy and the mean LDL-C value was 4 mmol/L (155 mg/dL) in the overall sample.
| All | Alive | Deceased | P | |
|---|---|---|---|---|
| ( n = 4930) | ( n = 4807) | ( n = 123) | ||
| Age (years) | 52 ± 10 | 52 ± 10 | 55 ± 11 | < 0.01 |
| Men | 2927 (59.4) | 2821 (58.7) | 106 (86.2) | < 0.01 |
| Current smoker | 1248 (25.3) | 1198 (24.9) | 50 (40.7) | < 0.01 |
| Body mass index (kg/m 2 ) | 26 ± 9 | 26 ± 9 | 25 ± 5 | 0.3 |
| Diabetes | 255 (5.2) | 237 (4.9) | 18 (14.6) | < 0.01 |
| Family history of premature CVD | 723 (14.7) | 708 (14.7) | 15 (12.2) | 0.62 |
| Systolic blood pressure (mmHg) | 136 ± 18 | 136 ± 17 | 146 ± 20 | < 0.01 |
| Diastolic blood pressure (mmHg) | 83 ± 9 | 82 ± 9 | 86 ± 10 | < 0.01 |
| Pulse pressure (mmHg) | 54 ± 13 | 53 ± 13 | 60 ± 16 | < 0.01 |
| Estimated GFR (mL/min/1.73 m 2 ) | 80 ± 17 | 80 ± 17 | 79 ± 18 | 0.5 |
| Antihypertensive drug treatment | 981 (19.9) | 952 (19.8) | 29 (23.6) | 0.3 |
| Statins | 1323 (26.8) | 1298 (27.0) | 25 (20.3) | 0.09 |
| Total cholesterol (mmol/L) | 6.17 ± 1.30 | 6.16 ± 1.30 | 6.46 ± 1.45 | 0.01 |
| LDL-C (mmol/L) a | 4.00 ± 1.19 | 3.99 ± 1.19 | 4.37 ± 1.21 | < 0.01 |
| HDL-C (mmol/L) | 1.41 ± 0.42 | 1.41 ± 0.42 | 1.30 ± 0.38 | < 0.01 |
| Triglycerides (mmol/L) | 1.75 ± 1.72 | 1.75 ± 1.73 | 1.79 ± 1.21 | 0.79 |
| LDL-C < 4 mmol/L (< 155 mg/dL) a | 2632 (54.6) | 2587 (55.0) | 45 (37.2) | < 0.01 |
| 4 ≤ LDL-C < 5 mmol/L (155 to < 190 mg/dL) a | 1367 (28.3) | 1318 (28.0) | 49 (40.5) | < 0.01 |
| 5 ≤ LDL-C < 6.5 mmol/L (190 to < 250 mg/dL) a | 684 (14.2) | 663 (14.1) | 21 (17.4) | 0.31 |
| 6.5 ≤ LDL-C < 8.5 mmol/L (250 to < 330 mg/dL) a | 123 (2.5) | 118 (2.5) | 5 (4.1) | 0.26 |
| LDL-C ≥ 8.5 mmol/L (≥ 330 mg/dL) a | 18 (0.4) | 17 (0.4) | 1 (0.8) | 0.41 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
