Dyslipidemia in the Young: From Genotype to Treatment Balder, Jan-Willem

Dyslipidemia in the Young: From Genotype to Treatment Balder, Jan-Willem

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date:

2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Balder, J-W. (2018). Dyslipidemia in the Young: From Genotype to Treatment. Rijksuniversiteit Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 29-06-2021

5

J.W. Balder

, P.J. Lansberg

, M.H. Hof

, A. Wiegman

, B.A. Hutten

, J.A. Kuivenhoven

Pediatric Lipid Reference Values in the General Population: the Dutch Lifelines Cohort Study

1. Department of Pediatrics, Section Molecular Genetics, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands.

2. Department of Vascular Medicine, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands.

3. Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands.

4. Department of Pediatrics, Emma Children’s Hospital, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands.

*Corresponding author: j.a.kuivenhoven@umcg.nl

Submitted to Journal of Clinical Lipidology J.W. Balder

, P.J. Lansberg

, M.H. Hof

, A. Wiegman

, B.A. Hutten 5

,

J.A. Kuivenhoven

Pediatric Lipid Reference Values in the General Population: the Dutch Lifelines Cohort Study

1. Department of Pediatrics, Section Molecular Genetics, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands.

2. Department of Vascular Medicine, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands.

3. Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands.

4. Department of Pediatrics, Emma Children’s Hospital, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands.

*Corresponding author: j.a.kuivenhoven@umcg.nl

Submitted to Journal of Clinical Lipidology

76

Abstract

Background

Atherosclerosis starts in childhood and its progression is influenced by lifelong low- density lipoprotein cholesterol (LDL-c) exposure, the so-called cholesterol burden. Early identification of children and adolescents with severely elevated LDL-c is thus of major clinical significance. This is especially true for children with familial hypercholesterolemia (FH), a frequent but undertreated genetic disorder. To identify children with possible FH, insight in the distribution of lipid levels in children is a prerequisite.

Objective

To provide health care professionals with contemporary age- and gender-based pediatric reference values for lipid and lipoprotein levels in order to be able to identify children with dyslipidemia, especially FH.

Methods

Lifelines is a large prospective population-based Dutch cohort study. Children from 8 till 18 years of age were included and fasting lipid levels were measured. Smoothed reference curves and percentiles (5

, 10

, 25

, 50

, 75

, 90

and 95

) were generated using the Generalized Additive Models for Location, Scale and Shape package in the statistical software R.

Results

A total of 8,071 children (3,823 boys and 4,248 girls) were included. In the overall cohort, we noted marked dynamic changes in lipid and lipoprotein levels over age, which in some cases were gender specific. Our data highlight a high and unexpected prevalence of severely elevated LDL-c (> 190 mg/dl) in both boys and girls.

Conclusions

Our cross-sectional data provide contemporary reference ranges for plasma lipids that can assist physicians in identifying children at increased risk of premature atherosclerosis, especially FH.

76

Abstract

Background

Atherosclerosis starts in childhood and its progression is influenced by lifelong low- density lipoprotein cholesterol (LDL-c) exposure, the so-called cholesterol burden. Early identification of children and adolescents with severely elevated LDL-c is thus of major clinical significance. This is especially true for children with familial hypercholesterolemia (FH), a frequent but undertreated genetic disorder. To identify children with possible FH, insight in the distribution of lipid levels in children is a prerequisite.

Objective

To provide health care professionals with contemporary age- and gender-based pediatric reference values for lipid and lipoprotein levels in order to be able to identify children with dyslipidemia, especially FH.

Methods

Lifelines is a large prospective population-based Dutch cohort study. Children from 8 till 18 years of age were included and fasting lipid levels were measured. Smoothed reference curves and percentiles (5

, 10

, 25

, 50

, 75

, 90

and 95

) were generated using the Generalized Additive Models for Location, Scale and Shape package in the statistical software R.

Results

A total of 8,071 children (3,823 boys and 4,248 girls) were included. In the overall cohort, we noted marked dynamic changes in lipid and lipoprotein levels over age, which in some cases were gender specific. Our data highlight a high and unexpected prevalence of severely elevated LDL-c (> 190 mg/dl) in both boys and girls.

Conclusions

Our cross-sectional data provide contemporary reference ranges for plasma lipids that can assist physicians in identifying children at increased risk of premature atherosclerosis, especially FH.

5

76 77

Introduction

Cardiovascular disease (CVD) represents a leading cause of death globally.

Most often men and women above 55 and 65 years of age respectively, are affected, but fatty streaks start developing at a very young age and the progression of atherosclerosis is positively associated with plasma low-density lipoprotein cholesterol (LDL-c) exposure.

This progression is accelerated in individuals with familial hypercholesterolemia (FH), a genetic disorder characterized by elevated LDL-c levels and premature CVD. Mutations in APOB, LDLR, and PCSK9 have been shown to cause FH. The prevalence of FH is estimated to be 1 per 200 – 250 individuals but this disease is severely underdiagnosed and undertreated.

Carriers of FH mutations suffer from increased cardiovascular risk, likely due to lifelong exposure to increased LDL-c levels.

In adults, and to a lesser extent in children, identification of FH affected individuals is restricted due to the overlap of LDL-c levels in both FH and non-FH individuals.

Therefore genetic testing can help in the clinical diagnosis of FH, and moreover in the screening of affected family members, known as cascade screening.

Childhood is the best period to discriminate between mutation-positive and mutation- negative hypercholesterolemia by measuring plasma LDL-c levels only.

Children with LDL-c levels twice > 190 mg/dl should be considered as having FH, while LDL-c levels twice

> 160 mg/dl in combination with a family history of hypercholesterolemia or premature CVD are highly suggestive of FH. Children of affected parent(s) with an LDL-c > 130 mg/dl are likely to be mutation carriers as well.

The aggressive progression of atherosclerosis in FH individuals can be effectively attenuated by means of lipid-lowering drugs, e.g. statins, ezetimibe and PCSK9-inhibition.

110

It has been shown that statin treatment in children was associated with normalization of carotid intima media thickness progression, a measure of subclinical atherosclerosis.

Such evidence has supported the initiation of statin treatment early in life (around 8 years) as an accepted strategy in clinical practice.

Therefore early identification (and possibly treatment) of children with FH is clinically relevant.

Identification of suspected FH children requires knowledge on the distribution of LDL-c levels in a pediatric population. To date, reference ranges are used from much older data or from small studies.

To provide such reference ranges, we used data of Lifelines, a prospective population- based cohort study which was initiated in 2006.

Using the same cohort, we recently reported that lipid levels in adults are strongly age- and gender-dependent, while the data of participants below the age of 18 years were not yet released.

For the current study, we have generated age- and gender-based reference values for lipid levels in children, aged 8 till 18 years. These data can help the identification of children at increased risk of atherosclerosis such as children with FH and assist cascade screening in families.

5

76 77

Introduction

Cardiovascular disease (CVD) represents a leading cause of death globally.

Most often men and women above 55 and 65 years of age respectively, are affected, but fatty streaks start developing at a very young age and the progression of atherosclerosis is positively associated with plasma low-density lipoprotein cholesterol (LDL-c) exposure.

This progression is accelerated in individuals with familial hypercholesterolemia (FH), a genetic disorder characterized by elevated LDL-c levels and premature CVD. Mutations in APOB, LDLR, and PCSK9 have been shown to cause FH. The prevalence of FH is estimated to be 1 per 200 – 250 individuals but this disease is severely underdiagnosed and undertreated.

Carriers of FH mutations suffer from increased cardiovascular risk, likely due to lifelong exposure to increased LDL-c levels.

In adults, and to a lesser extent in children, identification of FH affected individuals is restricted due to the overlap of LDL-c levels in both FH and non-FH individuals.

Therefore genetic testing can help in the clinical diagnosis of FH, and moreover in the screening of affected family members, known as cascade screening.

Childhood is the best period to discriminate between mutation-positive and mutation- negative hypercholesterolemia by measuring plasma LDL-c levels only.

Children with LDL-c levels twice > 190 mg/dl should be considered as having FH, while LDL-c levels twice

> 160 mg/dl in combination with a family history of hypercholesterolemia or premature CVD are highly suggestive of FH. Children of affected parent(s) with an LDL-c > 130 mg/dl are likely to be mutation carriers as well.

The aggressive progression of atherosclerosis in FH individuals can be effectively attenuated by means of lipid-lowering drugs, e.g. statins, ezetimibe and PCSK9-inhibition.

110

It has been shown that statin treatment in children was associated with normalization of carotid intima media thickness progression, a measure of subclinical atherosclerosis.

Such evidence has supported the initiation of statin treatment early in life (around 8 years) as an accepted strategy in clinical practice.

Therefore early identification (and possibly treatment) of children with FH is clinically relevant.

Identification of suspected FH children requires knowledge on the distribution of LDL-c levels in a pediatric population. To date, reference ranges are used from much older data or from small studies.

To provide such reference ranges, we used data of Lifelines, a prospective population- based cohort study which was initiated in 2006.

Using the same cohort, we recently reported that lipid levels in adults are strongly age- and gender-dependent, while the data of participants below the age of 18 years were not yet released.

For the current study, we have generated age- and gender-based reference values for lipid levels in children, aged 8 till 18 years. These data can help the identification of children at increased risk of atherosclerosis such as children with FH and assist cascade screening in families.

78

Methods

Study population

The study protocol was approved by Medical Ethical Committee of the University Medical Center Groningen (UMCG) in the Netherlands and all participants provided written informed consent. The rationale and design of Lifelines has been described previously.

In short, Lifelines is an ongoing prospective population-based cohort study. Between 2006 and 2013 inhabitants from the three northern provinces of the Netherlands (Groningen, Friesland and Drenthe) between 25 and 50 years of age were approached by their general practitioner to participate. Upon a positive response, relatives (first-degree family members, including children [≥ 6 months], partner and parents-in-law) were also invited. Individuals could also participate through self-registration. Of the 167,729 almost exclusively Caucasian participants, 14,801 are children. This multiple-generation design offers the unique opportunity to study the origins of multifactorial diseases. In total 85,000 (51%) participants are part of a two- generation family and 20,000 (12%) of a three-generation family.

For the present study, we provide cross-sectional population distributions of plasma lipid levels of children screened at the baseline visit. Blood sampling was only performed in children aged eight years or older. Children who were not fasted (defined as overnight fast) were excluded, as well as children with missing lipids measurements. In total 8,071 children (3,823 boys and 4,248 girls) were included. Supplementary Table 1 provides the number of children included for each year of age.

Questionnaires, physical examination and biomaterial collection

The parents of the children received questionnaires specifically suited for the child’s age. The questionnaires covered topics on lifestyle, health, nutrition and development. A physical examination was performed including anthropometry, blood pressure measurement (10 measurements during 10 minutes using a Dynamap registration) and pulmonary function tests. Fasting blood samples were drawn after an overnight fast. Fresh samples were transferred to the central laboratory of UMCG for routine clinical chemistry.

Cholesterol measurements

Total cholesterol, LDL-c, high-density lipoprotein cholesterol (HDL-c) and triglycerides were directly measured and were standardized against appropriate controls as described.

LDL-c levels were also calculated using the Friedewald formula,

when triglyceride levels did not exceed 400 mg/dl.

78

Methods

Study population

The study protocol was approved by Medical Ethical Committee of the University Medical Center Groningen (UMCG) in the Netherlands and all participants provided written informed consent. The rationale and design of Lifelines has been described previously.

In short, Lifelines is an ongoing prospective population-based cohort study. Between 2006 and 2013 inhabitants from the three northern provinces of the Netherlands (Groningen, Friesland and Drenthe) between 25 and 50 years of age were approached by their general practitioner to participate. Upon a positive response, relatives (first-degree family members, including children [≥ 6 months], partner and parents-in-law) were also invited. Individuals could also participate through self-registration. Of the 167,729 almost exclusively Caucasian participants, 14,801 are children. This multiple-generation design offers the unique opportunity to study the origins of multifactorial diseases. In total 85,000 (51%) participants are part of a two- generation family and 20,000 (12%) of a three-generation family.

For the present study, we provide cross-sectional population distributions of plasma lipid levels of children screened at the baseline visit. Blood sampling was only performed in children aged eight years or older. Children who were not fasted (defined as overnight fast) were excluded, as well as children with missing lipids measurements. In total 8,071 children (3,823 boys and 4,248 girls) were included. Supplementary Table 1 provides the number of children included for each year of age.

Questionnaires, physical examination and biomaterial collection

The parents of the children received questionnaires specifically suited for the child’s age. The questionnaires covered topics on lifestyle, health, nutrition and development. A physical examination was performed including anthropometry, blood pressure measurement (10 measurements during 10 minutes using a Dynamap registration) and pulmonary function tests. Fasting blood samples were drawn after an overnight fast. Fresh samples were transferred to the central laboratory of UMCG for routine clinical chemistry.

Cholesterol measurements

Total cholesterol, LDL-c, high-density lipoprotein cholesterol (HDL-c) and triglycerides were directly measured and were standardized against appropriate controls as described.

LDL-c levels were also calculated using the Friedewald formula,

when triglyceride levels did not exceed 400 mg/dl.

5

78 79

Statistics

Baseline characteristics that follow a normal distribution were reported as mean and standard deviation (SD). Baseline characteristics with a skewed distribution were reported as median and interquartile range (IQR).

Smoothed reference curves were generated using Generalized Additive Models for Location, Scale and Shape.

Let Y (t) be an outcome variable at age t. We used the Box- Cox-t power transformation with parameters μ (t) , σ (t), υ (t) to transform Y (t) to Z (t) with the following formula

:

The transformed variable Z (t) then follows a student t-distribution with τ (t) degrees of freedom. The relations between the age t and all four parameters were modeled with p-splines functions.

To compare our results to previously reported age- and gender-based percentiles, we also calculated age- and gender-based specific percentiles with corresponding age groups.

All analyses were performed stratified by sex and carried out using IBM SPSS Statistics, version 22.0 (Armonk, NY: IBM Corp.), or the statistical software R (version 3.2.2).

Results

Study population

Among the total of 167,729 individuals who participated in the Lifelines study, 14,801 (8.8%) were younger than 18 years. Of these, we excluded children who were younger than eight years of age because no blood was drawn at the baseline visit (n = 5,137), children with missing measurements (n = 1,103), and those without a fasting blood sample (n = 490). A total of 8,071 children were included in our study. Demographic and clinical characteristics of the study population are shown in Table 1. Mean age was, for both boys and girls, 12 years.

Furthermore, Table 1 shows that boys and girls have similar lipid profiles, blood pressure and glucose levels. Supplementary Table 1 shows the number of included children per year of age.

The associations between age and fasting lipid parameters are depicted in Figure 1 for boys and girls separately, using the 5

, 10

, 25

, 50

, 75

, 90

and 95

percentiles.

Supplementary Table 2 shows age- and gender-specific lipid percentiles, including the 5

, 10

, 50

, 90

and 95

percentiles. In the following sections we briefly describe the main findings for different components of the lipid profile.

5

78 79

Statistics

Baseline characteristics that follow a normal distribution were reported as mean and standard deviation (SD). Baseline characteristics with a skewed distribution were reported as median and interquartile range (IQR).

Smoothed reference curves were generated using Generalized Additive Models for Location, Scale and Shape.

Let Y (t) be an outcome variable at age t. We used the Box- Cox-t power transformation with parameters μ (t) , σ (t), υ (t) to transform Y (t) to Z (t) with the following formula

:

The transformed variable Z (t) then follows a student t-distribution with τ (t) degrees of freedom. The relations between the age t and all four parameters were modeled with p-splines functions.

To compare our results to previously reported age- and gender-based percentiles, we also calculated age- and gender-based specific percentiles with corresponding age groups.

All analyses were performed stratified by sex and carried out using IBM SPSS Statistics, version 22.0 (Armonk, NY: IBM Corp.), or the statistical software R (version 3.2.2).

Results

Study population

Among the total of 167,729 individuals who participated in the Lifelines study, 14,801 (8.8%) were younger than 18 years. Of these, we excluded children who were younger than eight years of age because no blood was drawn at the baseline visit (n = 5,137), children with missing measurements (n = 1,103), and those without a fasting blood sample (n = 490). A total of 8,071 children were included in our study. Demographic and clinical characteristics of the study population are shown in Table 1. Mean age was, for both boys and girls, 12 years.

Furthermore, Table 1 shows that boys and girls have similar lipid profiles, blood pressure and glucose levels. Supplementary Table 1 shows the number of included children per year of age.

The associations between age and fasting lipid parameters are depicted in Figure 1 for boys and girls separately, using the 5

, 10

, 25

, 50

, 75

, 90

and 95

percentiles.

Supplementary Table 2 shows age- and gender-specific lipid percentiles, including the 5

, 10

, 50

, 90

and 95

percentiles. In the following sections we briefly describe the main findings for different components of the lipid profile.

80

Age, gender and LDL-c levels

Boys aged 8 years presented with a median LDL-c of 89 mg/dl (95

percentile: 125 mg/dl).

Over the age range studied, LDL-c slightly decreased in boys as illustrated by both lower median and 95

percentile LDL-c levels between ages 13.0 and 16.9. Interestingly, at ages 17 and 18 LDL-c appears to increase marginally.

In girls, LDL-c levels appeared to be slightly higher compared to boys. As in the boys, a slight drop in LDL-c was noted at ages 11.0 to 14.9, thus a little earlier than in boys. The increase in LDL-c at later ages (15.0 – 17.9) was more pronounced in girls than boys. This is clearly illustrated at 17 years of age, when LDL-c is substantially higher in girls compared to boys for both median (82 mg/dl vs. 95 mg/dl) and the 95

percentile (126 mg/dl vs. 144 mg/dl).

In clinical practice LDL-c levels are often calculated using Friedewald formula (f-LDL-c).

Supplementary Figure 1 shows the absolute difference between f-LDL-c and direct measurement of LDL-c. Negative values indicate lower f-LDL-c in comparison to direct measurement of LDL-c and vice versa for positive values. Supplementary Figure 1 indicates that the population distribution of f-LDL-c is slightly lower in both boys and girls. Overall, f-LDL-c was approximately 6 mg/dl lower as compared to direct measurement of LDL-c.

Clearly, with increasing triglyceride levels, the discrepancy between direct measurement and f-LDL-c increases.

Age, gender and HDL-c levels

At 8 years of age, HDL-c levels are very similar in boys and girls, however, our cross- sectional data show markedly different HDL-c dynamics between genders with ageing.

HDL-c increases in boys aged 8.0 to 10.9 followed by a strong decrease between ages 11.0 and 14.9, and a subsequent stabilization. In girls, by contrast, HDL-c are generally slowly decreasing with age over the entire age range. At age 18, however, median, 75

, 90

and 95

percentiles of HDL-c are all higher in girls compared to boys.

80

Age, gender and LDL-c levels

Boys aged 8 years presented with a median LDL-c of 89 mg/dl (95

percentile: 125 mg/dl).

Over the age range studied, LDL-c slightly decreased in boys as illustrated by both lower median and 95

percentile LDL-c levels between ages 13.0 and 16.9. Interestingly, at ages 17 and 18 LDL-c appears to increase marginally.

In girls, LDL-c levels appeared to be slightly higher compared to boys. As in the boys, a slight drop in LDL-c was noted at ages 11.0 to 14.9, thus a little earlier than in boys. The increase in LDL-c at later ages (15.0 – 17.9) was more pronounced in girls than boys. This is clearly illustrated at 17 years of age, when LDL-c is substantially higher in girls compared to boys for both median (82 mg/dl vs. 95 mg/dl) and the 95

percentile (126 mg/dl vs. 144 mg/dl).

In clinical practice LDL-c levels are often calculated using Friedewald formula (f-LDL-c).

Supplementary Figure 1 shows the absolute difference between f-LDL-c and direct measurement of LDL-c. Negative values indicate lower f-LDL-c in comparison to direct measurement of LDL-c and vice versa for positive values. Supplementary Figure 1 indicates that the population distribution of f-LDL-c is slightly lower in both boys and girls. Overall, f-LDL-c was approximately 6 mg/dl lower as compared to direct measurement of LDL-c.

Clearly, with increasing triglyceride levels, the discrepancy between direct measurement and f-LDL-c increases.

Age, gender and HDL-c levels

At 8 years of age, HDL-c levels are very similar in boys and girls, however, our cross- sectional data show markedly different HDL-c dynamics between genders with ageing.

HDL-c increases in boys aged 8.0 to 10.9 followed by a strong decrease between ages 11.0 and 14.9, and a subsequent stabilization. In girls, by contrast, HDL-c are generally slowly decreasing with age over the entire age range. At age 18, however, median, 75

, 90

and 95

percentiles of HDL-c are all higher in girls compared to boys.

5

80 81

Figure 1 (continued on next page, legend follows)

5

80 81

Figure 1 (continued on next page, legend follows)

82

Figure 1. Age- and gender-based smoothed percentile curves (5^th-10^th-25^th-50^th-75^th-90^th-95^th) for total cholesterol, LDL-c, HDL-c and triglycerides for boys and girls separately.

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol.

Table 1. Demographics and clinical characteristics. Data is expressed as mean ± standard deviation but triglycerides as interquartile range.

Boys Girls

(n = 3,823) (n = 4,248)

Age (years) 12 ± 2.7 12 ± 2.8

Lipid profile (mg/dl)

Total cholesterol 155 ± 27 162 ± 27

LDL-c 89 ± 23 93 ± 23

HDL-c 58 ± 12 62 ± 12

Triglycerides 52 [40– 71] 62 [47 – 82]

SBP (mmHg) 107 ± 11 106 ± 11

DBP (mmHg) 59 ± 6 60 ± 6

Glucose (mmol/l) 4.7 ± 0.5 4.6 ± 0.5

HbA1c (%) 5.4 (0.3) 5.4 (0.3)

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure.

82

Figure 1. Age- and gender-based smoothed percentile curves (5^th-10^th-25^th-50^th-75^th-90^th-95^th) for total cholesterol, LDL-c, HDL-c and triglycerides for boys and girls separately.

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol.

Table 1. Demographics and clinical characteristics. Data is expressed as mean ± standard deviation but triglycerides as interquartile range.

Boys Girls

(n = 3,823) (n = 4,248)

Age (years) 12 ± 2.7 12 ± 2.8

Lipid profile (mg/dl)

Total cholesterol 155 ± 27 162 ± 27

LDL-c 89 ± 23 93 ± 23

HDL-c 58 ± 12 62 ± 12

Triglycerides 52 [40– 71] 62 [47 – 82]

SBP (mmHg) 107 ± 11 106 ± 11

DBP (mmHg) 59 ± 6 60 ± 6

Glucose (mmol/l) 4.7 ± 0.5 4.6 ± 0.5

HbA1c (%) 5.4 (0.3) 5.4 (0.3)

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure.

5

82 83

Age, gender and triglyceride levels

At 8 years of age triglyceride levels are lower in boys (median: 45 mg/dl; 95

percentile: 96 mg/dl) than in girls (median: 52 mg/dl; 95

percentile: 105 mg/dl). In both boys and girls, triglycerides levels increase with age over all percentiles studied. At the age of 18 years, median triglycerides were 47% higher in boys and 46% higher in girls compared to the boys and girls aged 8 years.

Extremes

In our study population, a considerable number of children presented with severe hypercholesterolemia: 375 (4.6%) children presented with LDL-c levels > 130 mg/dl. Eighty (1.0%) children showed LDL-c > 160 mg/dl, which is a strong indicator of having FH. Eighteen children (0.22%) presented with LDL-c levels > 190 mg/dl and could be considered as having familial hypercholesterolemia (Table 2).

Discussion

This study provides tables with age- and gender-based percentiles for lipid and lipoprotein levels in children and adolescents which can aid clinicians in the diagnosis of dyslipidemia, and in their decisions if additional diagnostic evaluations are indicated. This will allow to distinguish severe dyslipidemia, where pharmacological therapy might be indicated, from lipid abnormalities in need of lifestyle advice. We believe that the presented reference ranges are a prerequisite for effective screening and identification of children and adolescents with familial hypercholesterolemia (FH). Furthermore our data that were generated to screen for pediatric hypercholesterolemia (LDL-c > 190 mg/dl, indicative of FH), revealed a prevalence of 1:450, which illustrates the need for continued alertness to identify as well as treat children and adolescents with FH. Unfortunately, however, the use of our data may be limited primarily to children and adolescents of Caucasian descent.

Importance of lipid reference values in clinical practice

Several studies have shown that young children with increased total cholesterol levels, e.g.

above the 90

percentile, will maintain their percentile ranking over time.

This is why identifying and targeting children and adolescents with FH, is an important step in reducing cardiovascular events. Cost-effective interventions using low cost generic statins combined with cholesterol absorption inhibitors can produce substantial reductions in LDL-c levels.

Current guidelines advocate initiating statin treatment in boys and girls with FH from the age of 8 years onwards.

Although long-term data on safety and cardiovascular outcomes for statin treated children is limited, side-effects are rare

and a 10-year follow-up study

5

82 83

Age, gender and triglyceride levels

At 8 years of age triglyceride levels are lower in boys (median: 45 mg/dl; 95

percentile: 96 mg/dl) than in girls (median: 52 mg/dl; 95

percentile: 105 mg/dl). In both boys and girls, triglycerides levels increase with age over all percentiles studied. At the age of 18 years, median triglycerides were 47% higher in boys and 46% higher in girls compared to the boys and girls aged 8 years.

Extremes

In our study population, a considerable number of children presented with severe hypercholesterolemia: 375 (4.6%) children presented with LDL-c levels > 130 mg/dl. Eighty (1.0%) children showed LDL-c > 160 mg/dl, which is a strong indicator of having FH. Eighteen children (0.22%) presented with LDL-c levels > 190 mg/dl and could be considered as having familial hypercholesterolemia (Table 2).

Discussion

This study provides tables with age- and gender-based percentiles for lipid and lipoprotein levels in children and adolescents which can aid clinicians in the diagnosis of dyslipidemia, and in their decisions if additional diagnostic evaluations are indicated. This will allow to distinguish severe dyslipidemia, where pharmacological therapy might be indicated, from lipid abnormalities in need of lifestyle advice. We believe that the presented reference ranges are a prerequisite for effective screening and identification of children and adolescents with familial hypercholesterolemia (FH). Furthermore our data that were generated to screen for pediatric hypercholesterolemia (LDL-c > 190 mg/dl, indicative of FH), revealed a prevalence of 1:450, which illustrates the need for continued alertness to identify as well as treat children and adolescents with FH. Unfortunately, however, the use of our data may be limited primarily to children and adolescents of Caucasian descent.

Importance of lipid reference values in clinical practice

Several studies have shown that young children with increased total cholesterol levels, e.g.

above the 90

percentile, will maintain their percentile ranking over time.

This is why identifying and targeting children and adolescents with FH, is an important step in reducing cardiovascular events. Cost-effective interventions using low cost generic statins combined with cholesterol absorption inhibitors can produce substantial reductions in LDL-c levels.

Current guidelines advocate initiating statin treatment in boys and girls with FH from the age of 8 years onwards.

Although long-term data on safety and cardiovascular outcomes for statin treated children is limited, side-effects are rare

and a 10-year follow-up study

84

showed normalization of progression of carotid intima media thickness in statin-treated FH children.

The real challenge is how to find children with FH. The most commonly used diagnostic criteria for FH (i.e. Dutch Lipid Clinic Network and Simon Broome criteria) are importantly not applicable to children. Currently an LDL-c level threshold of 190 mg/dl is used as diagnostic criterion while a threshold of 130 mg/dl in case of affected family members is advocated by experts.

Universal pediatric screening of FH could be easily and effectively performed when incorporated in infant health check-ups such as immunization programs.

This has been proven to be a highly effective approach, especially if reverse cascade screening, to identify additional affected family members, is integrated.

In line with cross-sectional studies, we show that lipid distributions change with age.

118, 119

Instead of using a strict absolute LDL-c cutoff level to identify children and adolescents with abnormal lipid levels, we propose using age- and gender-specific cutoff percentiles. In our opinion, children with LDL-c levels exceeding the 90

percentile should be monitored and lifestyle interventions should be considered.

When LDL-c levels exceed the 95

percentile, a molecular diagnosis should be pursued, especially in case of familial presence of hypercholesterolemia or premature CVD.

Comparison to LRC Prevalence study

The LRC Prevalence study conducted in the USA in 1970s

has provided clinicians and researchers with cross-sectional lipid reference values.

After matching age groups, we compared the age- and gender-based percentiles from LRC Prevalence study to the Lifelines study (Figure 2). LDL-c levels were lower in boys (~10 mg/dl) of the Lifelines cohort during adolescence while this discrepancy was not observed in girls. HDL-c levels were higher in both boys and girls in Lifelines compared to the LRC study while triglycerides were slightly lower. Differences in methodology, population characteristics and time of inclusion (difference of ~30 year) could possibly explain discrepancies. The US National Health and Nutrition Examination Surveys (NHANES) have showed an initial decline of total cholesterol levels in children and young adults (aged 4 – 19 years) between 1966 – 1970 and 1988 – 1994.

The detrimental increase in overweight and obesity in children over the past decades,

which is associated with dyslipidemia,

may have reversed the initial lowering of LDL-c levels observed at that time.

84

showed normalization of progression of carotid intima media thickness in statin-treated FH children.

The real challenge is how to find children with FH. The most commonly used diagnostic criteria for FH (i.e. Dutch Lipid Clinic Network and Simon Broome criteria) are importantly not applicable to children. Currently an LDL-c level threshold of 190 mg/dl is used as diagnostic criterion while a threshold of 130 mg/dl in case of affected family members is advocated by experts.

Universal pediatric screening of FH could be easily and effectively performed when incorporated in infant health check-ups such as immunization programs.

This has been proven to be a highly effective approach, especially if reverse cascade screening, to identify additional affected family members, is integrated.

In line with cross-sectional studies, we show that lipid distributions change with age.

118, 119

Instead of using a strict absolute LDL-c cutoff level to identify children and adolescents with abnormal lipid levels, we propose using age- and gender-specific cutoff percentiles. In our opinion, children with LDL-c levels exceeding the 90

percentile should be monitored and lifestyle interventions should be considered.

When LDL-c levels exceed the 95

percentile, a molecular diagnosis should be pursued, especially in case of familial presence of hypercholesterolemia or premature CVD.

Comparison to LRC Prevalence study

The LRC Prevalence study conducted in the USA in 1970s

has provided clinicians and researchers with cross-sectional lipid reference values.

After matching age groups, we compared the age- and gender-based percentiles from LRC Prevalence study to the Lifelines study (Figure 2). LDL-c levels were lower in boys (~10 mg/dl) of the Lifelines cohort during adolescence while this discrepancy was not observed in girls. HDL-c levels were higher in both boys and girls in Lifelines compared to the LRC study while triglycerides were slightly lower. Differences in methodology, population characteristics and time of inclusion (difference of ~30 year) could possibly explain discrepancies. The US National Health and Nutrition Examination Surveys (NHANES) have showed an initial decline of total cholesterol levels in children and young adults (aged 4 – 19 years) between 1966 – 1970 and 1988 – 1994.

The detrimental increase in overweight and obesity in children over the past decades,

which is associated with dyslipidemia,

may have reversed the initial lowering of LDL-c levels observed at that time.

5

84 85

Figure 2 (continued on next page, legend follows)

5

84 85

Figure 2 (continued on next page, legend follows)

86

Figure 2. Comparison of total cholesterol, LDL-c, HDL-c and triglycerides between Lifelines (dotted line) and LRC Prevalence study (solid line). The lines correspond to the 5^th, median and 95^th percentile.

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; LRC, Lipid Research Clinics.

Table 2. Prevalence of dyslipidemia in Lifelines, compared to Germany and United States.

Prevalence in

Lifelines - n (%) Prevalence in

Germany¹¹⁹ Prevalence

in US^{120, 122}

Total cholesterol > 200 mg/dl 412 (5.1%) 7.8 % 8%

LDL-c > 130 mg/dl 375 (4.6%) 6.1% 7%

LDL-c > 160 mg/dl 80 (1.0%) - -

LDL-c > 190 mg/dl 18 (0.22%) 0.23% -

HDL-c < 40 mg/dl 362 (4.5%) 8.0% 13 – 15%

Triglycerides > 133 mg/dl 282 (3.5%) 11.7% 12%

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol.

Frequency of pediatric hypercholesterolemia

In our study population, 4.6% of the children presented with LDL-c levels > 130 mg/dl. In the United States and Germany higher percentages were reported, 7.0 and 6.1% respectively (Table 2). The lower frequency of hypercholesterolemia in our cohort may be explained by the notion that we were not able to account for possible effects of medication, diet or secondary causes of dyslipidemia.

86

Figure 2. Comparison of total cholesterol, LDL-c, HDL-c and triglycerides between Lifelines (dotted line) and LRC Prevalence study (solid line). The lines correspond to the 5^th, median and 95^th percentile.

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; LRC, Lipid Research Clinics.

Table 2. Prevalence of dyslipidemia in Lifelines, compared to Germany and United States.

Prevalence in

Lifelines - n (%) Prevalence in

Germany¹¹⁹ Prevalence

in US^{120, 122}

Total cholesterol > 200 mg/dl 412 (5.1%) 7.8 % 8%

LDL-c > 130 mg/dl 375 (4.6%) 6.1% 7%

LDL-c > 160 mg/dl 80 (1.0%) - -

LDL-c > 190 mg/dl 18 (0.22%) 0.23% -

HDL-c < 40 mg/dl 362 (4.5%) 8.0% 13 – 15%

Triglycerides > 133 mg/dl 282 (3.5%) 11.7% 12%

Abbreviations: LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol.

Frequency of pediatric hypercholesterolemia

In our study population, 4.6% of the children presented with LDL-c levels > 130 mg/dl. In the United States and Germany higher percentages were reported, 7.0 and 6.1% respectively (Table 2). The lower frequency of hypercholesterolemia in our cohort may be explained by the notion that we were not able to account for possible effects of medication, diet or secondary causes of dyslipidemia.

5

86 87

Conclusions

Our study provides contemporary age- and gender based reference values for plasma lipid and lipoprotein levels in children and adolescents. We expect that these lipid and lipoprotein values translated into age- and gender-specific percentiles, can serve clinicians as a simple, effective and reliable instrument to help identifying children and adolescents with dyslipidemia, especially FH.

Acknowledgments

We would like to thank all participants of the Lifelines study. We are furthermore indebted to Prof. J.J.P. Kastelein for critical reading of the manuscript.

Conflicts of interest and source of funding

Conflicts of interest: none. This work was by the Netherlands CardioVascular Research Initiative (Genius CVON2011-2016; and Genius2 CVON2017-2020) and the European Union (Transcard FP7-603091-2). JAK is Established Investigator of the Netherlands Heart Foundation (2015T068).

5

86 87

Conclusions

Our study provides contemporary age- and gender based reference values for plasma lipid and lipoprotein levels in children and adolescents. We expect that these lipid and lipoprotein values translated into age- and gender-specific percentiles, can serve clinicians as a simple, effective and reliable instrument to help identifying children and adolescents with dyslipidemia, especially FH.

Acknowledgments

We would like to thank all participants of the Lifelines study. We are furthermore indebted to Prof. J.J.P. Kastelein for critical reading of the manuscript.

Conflicts of interest and source of funding

Conflicts of interest: none. This work was by the Netherlands CardioVascular Research Initiative (Genius CVON2011-2016; and Genius2 CVON2017-2020) and the European Union (Transcard FP7-603091-2). JAK is Established Investigator of the Netherlands Heart Foundation (2015T068).

88

Supplemental material

Supplementary Figure 1. The absolute difference between Friedewald LDL-c and direct measurement of LDL-c are presented by triglyceride strata. The median, 5^th and 95^th percentiles are presented for boys (black) and girls (white) separately. Overall, direct measurement was slightly higher compared to calculated LDL-c.

Supplementary Table 1. Number of included children.

Age (years) Gender Total

Boys Girls

8 446 443 889

9 392 382 774

10 458 482 940

11 455 470 925

12 399 461 860

13 439 435 874

14 372 417 789

15 351 419 770

16 287 421 708

17 224 318 542

88

Supplemental material

Supplementary Figure 1. The absolute difference between Friedewald LDL-c and direct measurement of LDL-c are presented by triglyceride strata. The median, 5^th and 95^th percentiles are presented for boys (black) and girls (white) separately. Overall, direct measurement was slightly higher compared to calculated LDL-c.

Supplementary Table 1. Number of included children.

Age (years) Gender Total

Boys Girls

8 446 443 889

9 392 382 774

10 458 482 940

11 455 470 925

12 399 461 860

13 439 435 874

14 372 417 789

15 351 419 770

16 287 421 708

17 224 318 542

5

88 89

Supplementary Table 2. Age- and gender specific lipid percentiles.

Boys Girls

Age Percentile Age Percentile

5 10 50 90 95 5 10 50 90 95

Total cholesterol

8 120 128 155 185 195 8 126 133 159 192 203

9 121 130 157 188 199 9 127 135 162 195 206

10 123 131 159 191 202 10 128 135 162 196 208

11 123 131 159 193 205 11 126 134 161 195 207

12 121 130 158 192 205 12 124 131 159 193 205

13 117 125 154 188 200 13 122 129 157 191 203

14 112 119 147 181 193 14 120 128 155 190 202

15 108 115 143 176 188 15 120 128 156 192 204

16 106 114 142 176 187 16 121 129 159 196 208

17 108 116 145 180 192 17 123 131 162 201 215

18 109 118 148 185 198 18 125 134 167 208 222

Low-density lipoprotein cholesterol

8 59 66 89 115 125 8 63 69 92 121 131

9 59 66 89 117 127 9 64 70 94 124 134

10 58 65 89 118 130 10 63 70 94 125 136

11 58 65 90 120 131 11 61 68 93 124 135

12 57 64 89 120 131 12 59 66 90 121 132

13 55 61 86 117 128 13 57 63 88 119 130

14 52 58 82 113 125 14 55 62 87 118 128

15 51 57 80 111 122 15 55 62 88 121 132

16 50 56 80 111 123 16 56 63 91 126 137

17 51 58 82 114 126 17 58 65 95 131 144

18 53 60 85 119 132 18 59 67 99 138 152

High-density lipoprotein cholesterol

8 43 46 61 78 84 8 43 47 61 77 83

9 44 48 62 80 86 9 42 46 61 78 84

5

88 89

Supplementary Table 2. Age- and gender specific lipid percentiles.

Boys Girls

Age Percentile Age Percentile

5 10 50 90 95 5 10 50 90 95

Total cholesterol

8 120 128 155 185 195 8 126 133 159 192 203

9 121 130 157 188 199 9 127 135 162 195 206

10 123 131 159 191 202 10 128 135 162 196 208

11 123 131 159 193 205 11 126 134 161 195 207

12 121 130 158 192 205 12 124 131 159 193 205

13 117 125 154 188 200 13 122 129 157 191 203

14 112 119 147 181 193 14 120 128 155 190 202

15 108 115 143 176 188 15 120 128 156 192 204

16 106 114 142 176 187 16 121 129 159 196 208

17 108 116 145 180 192 17 123 131 162 201 215

18 109 118 148 185 198 18 125 134 167 208 222

Low-density lipoprotein cholesterol

8 59 66 89 115 125 8 63 69 92 121 131

9 59 66 89 117 127 9 64 70 94 124 134

10 58 65 89 118 130 10 63 70 94 125 136

11 58 65 90 120 131 11 61 68 93 124 135

12 57 64 89 120 131 12 59 66 90 121 132

13 55 61 86 117 128 13 57 63 88 119 130

14 52 58 82 113 125 14 55 62 87 118 128

15 51 57 80 111 122 15 55 62 88 121 132

16 50 56 80 111 123 16 56 63 91 126 137

17 51 58 82 114 126 17 58 65 95 131 144

18 53 60 85 119 132 18 59 67 99 138 152

High-density lipoprotein cholesterol

8 43 46 61 78 84 8 43 47 61 77 83

9 44 48 62 80 86 9 42 46 61 78 84

90

10 44 48 63 82 87 10 42 46 61 78 84

11 44 48 63 82 88 11 42 46 61 78 84

12 44 47 62 80 86 12 42 46 60 77 83

13 42 46 60 77 83 13 42 46 60 77 82

14 39 43 56 73 78 14 42 46 60 76 82

15 38 41 54 69 74 15 41 45 59 75 81

16 37 41 53 68 73 16 40 43 58 75 80

17 37 40 53 68 73 17 39 42 56 74 80

18 37 40 52 67 72 18 38 41 56 73 80

Triglycerides

8 25 28 45 80 96 8 29 33 52 89 105

9 26 29 47 81 99 9 30 34 53 92 110

10 26 29 48 84 102 10 31 35 56 97 117

11 27 30 49 87 105 11 33 37 58 104 125

12 27 31 50 89 109 12 34 38 61 107 129

13 28 33 52 94 114 13 35 39 62 108 129

14 30 34 55 98 120 14 35 39 63 107 128

15 31 35 58 103 125 15 35 40 64 108 128

16 33 37 60 107 130 16 37 42 67 112 132

17 35 39 64 112 136 17 39 45 71 119 140

18 36 42 66 118 143 18 42 48 76 126 149

Boys Girls

Age Percentile Age Percentile

5 10 50 90 95 5 10 50 90 95

Supplementary Table 2. Continued.

90

10 44 48 63 82 87 10 42 46 61 78 84

11 44 48 63 82 88 11 42 46 61 78 84

12 44 47 62 80 86 12 42 46 60 77 83

13 42 46 60 77 83 13 42 46 60 77 82

14 39 43 56 73 78 14 42 46 60 76 82

15 38 41 54 69 74 15 41 45 59 75 81

16 37 41 53 68 73 16 40 43 58 75 80

17 37 40 53 68 73 17 39 42 56 74 80

18 37 40 52 67 72 18 38 41 56 73 80

Triglycerides

8 25 28 45 80 96 8 29 33 52 89 105

9 26 29 47 81 99 9 30 34 53 92 110

10 26 29 48 84 102 10 31 35 56 97 117

11 27 30 49 87 105 11 33 37 58 104 125

12 27 31 50 89 109 12 34 38 61 107 129

13 28 33 52 94 114 13 35 39 62 108 129

14 30 34 55 98 120 14 35 39 63 107 128

15 31 35 58 103 125 15 35 40 64 108 128

16 33 37 60 107 130 16 37 42 67 112 132

17 35 39 64 112 136 17 39 45 71 119 140

18 36 42 66 118 143 18 42 48 76 126 149

Boys Girls

Age Percentile Age Percentile

5 10 50 90 95 5 10 50 90 95

Supplementary Table 2. Continued.

5

90 91

5

90 91