The study protocol was approved by the Regional Committee for Medical Health Research Ethics (#S-08337a/S-07455a) and by the Norwegian Data Inspectorate (#08/3914). A written informed consent was obtained from each patient included in the study and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki.
Collection of data from three surveys to estimate the prevalence of sHTG
The current data on prevalence originates from (i) the COhorts of NORway (CONOR; n = 172,956); (ii) the 3 Counties Study (n = 92,080); and (iii) the 40 Years Survey (n = 416,954). The data from these three surveys have in the present study been merged (n = 681,990) to generate data of the Norwegian population.
CONOR covers 10 population-based studies during 1994-2003 from different parts of Norway, both urban and rural regions. Some studies invited all residents above a specific age (for example all above 19 years in the HUNT 2 study), whereas others invited all subjects in selected age groups (for example all 30-, 40-, 45-, 60 and 75 years in the OPPHED- and TROFINN studies). The overall participation rate was 60% and varied between the studies. The participants (aged 20-103 years) filled in a questionnaire covering physical activity, history of previous cardiovascular disease (CVD), diabetes and use of medicine. A non-fasting blood sample was collected. Body mass index (BMI) and systolic blood pressure were also measured. For subjects who participated more than once we used the last measured TG values. Details of the research tools and the measurements of blood chemistry have been described elsewhere [4].
In the 3 Counties Study (covering the counties Oppland in Eastern part of Norway; Sogn og Fjordane in Western part of Norway; and Finnmark in Northern part of Norway; during 1974-1988) all residents aged 35-49 years at the study start in each county were invited to the first screening. The study had three visits in each county. The same birth cohorts and in some counties broader age groups, were invited to the second and third screening in each county. Altogether about 93,000 men and women participated in at least one screening. The participation rate varied from 78% to 90%. The participants were screened for CVD risk factors and a non-fasting blood sample was collected. We have included the last measured TG value for each participant in the present study. Details of the research tools and the measurements of blood chemistry have been described elsewhere [5, 6].
The 40 Years Survey was conducted from 1985 to 1999 and consisted of 429,245 participants from 18 counties. All inhabitants aged 40-42 years were invited to the survey whereas in some counties and municipalities broader age groups were included. The participation rate varied from 69% in Nord-Trøndelag county in 1992 to 46% in Østfold- and Aust-Agder counties in 1999. A non-fasting blood sample was collected. A few subjects participated more than once, and for them the last TG value was used in the present study. Details of the research tools and the measurements of blood chemistry have been described elsewhere [7].
From each study, we included all participants who had obtained measurements of serum triglycerides and who had answered the questionnaires.
Collection of data from patients with sHTG for detailed analyses
All individuals >18 years treated for sHTG at the Lipid Clinic from January 1st 2002 to December 31st 2007 were identified by searching the medical records for the diagnoses (ICD-10 classification): pure hypertriglyceridemia (E78.1), mixed hyperlipidemia (E78.2) and hyperchylomicronemia (E78.3). To ensure that the hypertriglyceridemia was not transient, only patients with long-lasting sHTG was included, i.e. they were diagnosed with sHTG both at the time of referral and at two consecutive time points 2-6 months apart. One hundred and twelve patients fulfilled the inclusion criteria and were invited by letter. Sixty-five (58%) patients returned a signed informed consent and were included in the study and came for a baseline visit. Inclusion and exclusion criteria are listed in the Additional file 1: Table S1. Data regarding age, diagnoses, clinical findings, laboratory measures, and use of medication were collected from their medical records. All laboratory values were collected after at least 12 h of fasting and analysed at the laboratory at Oslo University Hospital, using standard procedures. Most patients were controlled after 0.5, 1, 2 and 3 years (± 3 months) after the baseline visit. The genes lipoprotein lipase (LPL), apolipoprotein C-II (APOC2), apolipoprotein A-V (APOA5), glycosylphosphatidylinositol-anchored high density lipoprotein-binding protein 1 (GPIHBP1) and lipase maturation factor 1 (LMF1) were sequenced at the Unit for Cardiac and Cardiovascular Genetics, Department of Medical Genetics, Oslo University Hospital, and apolipoprotein E was genotyped. Information regarding diet and lifestyle were collected from the validated food questionnaire SmartDiet [8].
RNA expression of lipid-related genes in peripheral blood mononuclear cells
Six male patients with hypertriglyceridemia and nine male control subjects were recruited from employees of the Lipid Clinic and the University of Oslo, respectively, to this sub-study.
We analyzed a range of lipid-related genes (Additional file 1: Table S2) in peripheral blood mononuclear cells (PBMC) obtained from heparinized blood by gradient centrifugation in Isopaque-Ficoll (Lymphoprep, Nycomed, Oslo, Norway). The assay-numbers (hs-numbers) of the specific genes chosen for the analysis are included (Additional file 1: Table S2). PBMC pellets for mRNA expression analyses were immediately frozen and stored at −80 °C. Total RNA was isolated from all PBMC samples using RNeasy mini kit (Qiagen, Hilden, Germany) with lysis buffer containing β-mercaptoethanol, and RNase-Free DNase (Qiagen) and stored at −80 °C. RNA quantity and -quality measurements were performed using the ND 1000 Spectrophotometer (Saveen Werner, Carlson Circle Tampa, FL) and Agilent Bioanalyser (Agilent Technologies, Santa Clara, CA), respectively. Four hundred ng RNA from all samples were reverse-transcribed using High Capacity RNA-to-cDNA Kit (Applied Biosystems, Foster City, CA). Quantitative real-time polymerase chain reaction (RT-qPCR) was performed using Custom TaqMan Array micro Fluidic cards (Applied Biosystems). The endogenous control genes glucuronidase β (GUSβ) and TATA box binding protein (TBP) were used for normalization. The relative mRNA level for each transcript was calculated by the ΔΔ cycle threshold method [9].
Statistical analyses
The patient data were analysed with the Statistical Package for the Social Sciences (SPSS), version 16.0. Results from non-parametric statistical analyses are reported as median and 95% confidence interval (95% CI). Results from variables that showed a normal distribution are reported as mean and standard deviation (SD). For comparisons of two groups of individuals, the Mann-Whitney U test was used. Significance was assumed for p < 0.05. The prevalence data were analyzed with SPSS version 19.