Recruitment process and sample size

The Study was approved by the Institutional Ethic's Committee of the University of Lausanne and recruitment took place in the city of Lausanne in Switzerland, a town of 117,161 inhabitants, of which 79,420 are of a Swiss nationality[6].

The complete list of the Lausanne inhabitants aged 35–75 years (n = 56,694 in 2003) was provided by the population register of the city and served to sample the participants to the study. All subjects living in the city of Lausanne in 2003 for more than 90 days have their name included in this register. The register had information on age and gender but no information regarding ethnicity or country of origin.

Given the allele frequency of most SNPs on the Affymetrix chip and the genetic effect sizes expected for complex diseases (i.e. OR 1.3 and above), we chose to include ~6000 individuals, so as to have enough power (~80%) to detect genetic associations for diseases with a prevalence of ~15%.

A simple, non-stratified random selection of 19,830 subjects, corresponding to 35% of the source population, was drawn using STATA v9.1 software (Stata Corp, College Station, USA), and a letter inviting the addressee to participate in the study was sent to these individuals.

Subjects who volunteered to participate were contacted by phone within 14 days by one of the staff members to set up an appointment. Subjects who didn't answer were sent a second invitation letter. If no answer was obtained, they were contacted by phone. Subjects were considered as non-participants if they refused to participate and as non-responders if contact couldn't be made after two successive letters and three successive phone calls. Individuals who didn't live in Lausanne any longer, who were dead or who didn't meet the age criteria were considered as non-eligible. Recruitment began in June 2003 and ended in May 2006.

The sampling procedure is summarized in Figure 1. Of the initial 19,830 subjects sampled, 54 subjects were considered as non-eligible before contact and 15,109 (76%) responses were obtained. A total of 4667 subjects who did not respond were considered as non-responders. Among responders, 6,189 (41%) subjects refused to participate in the study and 799 (5%) were considered as non-eligible. Among the latter, 53% moved to a different city, 32% were out of the age range and 11% were reported to be deceased. The sample of 8,121 subjects who agreed to participate represented 41% of the initially sampled population, 54% of all responders and 57% of all eligible responders.

Inclusion criteria

The following inclusion criteria were applied: a) written informed consent; b) age 35–75 years and c) Caucasian origin. Caucasian origin was defined as having both parents and grandparents born in a restricted list of countries (available from the authors). No other exclusion criteria were applied. Since ethnicity could only be established during the clinic visit, some participants were assessed but were not included in the CoLaus study. Their data will be described elsewhere.

Assessment process

Participants were asked to attend the outpatient clinic at the Centre Hospitalier Universitaire Vaudois (CHUV) in the morning after an overnight fast. They had to take their medication as usual. Data were collected by trained field interviewers in a single visit lasting about 60 minutes. Informed consent was obtained from participants upon their arrival at the study clinic. The first questionnaire mailed with the appointment's letter and completed by the participant prior to the morning visit was then quickly reviewed and a second questionnaire was applied by interview prior to clinical measurements and blood collection.

Questionnaire data

The first set of questionnaires recorded information on demographic data, socio-economic and marital status, and several lifestyle factors, namely tobacco, alcohol and caffeine consumption, physical activity and mood. Data on smoking included the previous and current smoking status as well as the amount of tobacco smoked (number of cigarettes, cigarillos, cigars or pipes), age of beginning and end (for former smokers). Similarly, data on alcohol consumption included the past and current drinking status as well as the number of alcoholic beverage units (wine, beer and spirits) consumed over the week preceding the interview. Caffeine consumption was assessed by the number of caffeine-containing beverages consumed per day. Personal history of overweight and/or obesity and birth weight were also collected. Finally, the 12-item General Health Questionnaire (GHQ12) [7] was applied in order to screen for the presence of non-psychotic psychiatric disorders.

The second questionnaire, administered during a face-to-face meeting with the recruiter, focused on personal and family history of disease and CV risk factors. Subjects were asked which disease(s) they or their family had presented. When a positive answer was given, further information regarding age of occurrence and number of family members affected was collected. When appropriate, death of parents was recorded with age and cause of death. Regarding blood pressure (BP) status, subjects indicated if they had been diagnosed with hypertension and subsequently if they had been, or were being treated currently for this condition. BP levels before the beginning of treatment were sought for and the names of the anti-hypertensive drugs that had been prescribed were collected. In case the anti-hypertensive regimen had been modified, the duration and the reason for changing were also recorded. Personal medicines, including prescription and self-prescribed drugs, vitamin and mineral supplements, homeopathy or natural remedies were collected, together with their main indications. In women, further data regarding reproductive and obstetrical history, oral contraception and hormonal replacement therapy was collected. Finally, an additional frailty questionnaire (for subjects aged over 50 years) and the Mini-Mental State Evaluation (MMSE – for subjects aged over 65 years) were further administered [8].

Clinical data

Body weight and height were measured with participants standing without shoes in light indoor clothes. Body weight was measured in kilograms to the nearest 0.1 kg using a Seca^® scale (Hamburg, Germany), which was calibrated regularly. Height was measured to the nearest 5 mm using a Seca^® height gauge (Hamburg, Germany). Body mass index (BMI) was defined as weight/height². Obesity was defined as BMI ≥ 30 kg/m² and overweight as BMI ≥ 25 kg/m² and < 30 kg/m² .

BP and heart rate were measured thrice on the left arm, with an appropriately sized cuff, after at least 10 minute rest in the seated position using an Omron^® HEM-907 automated oscillometric sphygmomanometer (Matsusaka, Japan) [9]. The average of the last two measurements was used for analyses. Hypertension was defined as a systolic BP (SBP) ≥ 140 mm Hg and/or a diastolic BP (DBP) ≥ 90 mm Hg during the visit and/or presence of anti-hypertensive drug treatment and was considered as known if the subject was aware of this condition.

In addition, waist and hip circumferences were measured as recommended [10] and fat and fat-free mass were assessed by electrical bioimpedance [11] using the Bodystat^® 1500 analyzer (Isle of Man, British Isles). Finally, baldness and its age of onset were assessed in men using the Hamilton scale [12].

Sub-studies

In addition, sub-studies designed to assess the psychiatric characteristics of this population as well as functional CV measurements, were nested onto this study, and will be described separately.

Biological data

Venous blood samples (50 ml) were drawn after an overnight fast, and most clinical chemistry assays were performed by the CHUV Clinical Laboratory on fresh blood samples whereas Pathway Diagnostics (Los Angeles, CA) measured adiponectin, leptin and insulin [See additional file 1]. Additional aliquots were stored at -80°C.

LDL-cholesterol was calculated with the Friedewald formula only if triglycerides <4.6 mmol/l. Low HDL cholesterol level was defined as <1.0 mmol/L; high HDL cholesterol as ≥ 1.6 mmol/L; high LDL cholesterol was defined as ≥ 4.1 mmol/L and high triglyceride level was defined as ≥ 2.2 mmol/L [13]. In our analysis, dyslipidemia was defined as low HDL cholesterol and/or high triglyceride and/or LDL cholesterol ≥ 4.1 mmol/L or ≥ 2.6 mmol/L in presence of self-reported history of myocardial infarction, stroke, coronary artery disease or diabetes.

Diabetes was defined as fasting plasma glucose ≥ 7.0 mmol/L and/or presence of oral hypoglycaemic or insulin treatment. Type 2 diabetes mellitus (T2DM) was defined in case of diabetes without self-reported Type 1 DM. Diabetes was considered as known if the subject was aware of this condition. Impaired fasting glucose (IFG) was defined as fasting plasma glucose between 6.1 and 6.9 mmol/L without anti-diabetic treatment [14].

A urine sample was collected for the assessment of creatinine and albumin and the albumin-to-creatinine ratio was calculated. Microalbuminuria was defined as a value of albumin-to-creatinine ratio above 30 mg/g.

Genotyping

Nuclear DNA was extracted from whole blood for whole genome scan analysis and genotyping was performed using the Affimetrix 500 K SNP chip, as recommended by the manufacturer. Subjects consented for the genetic data to be used for the study of cardiovascular risk factors, and associated diseases including mood disorders.

Data management, security and quality control

Data were entered into a secured, internet-based database. The database was designed to confirm the validity of the identification codes, establish the completeness of the information keyed in and to perform basic data checks. All discrepancies were recorded in the case report form kept in a locked room. Each modification of the data was automatically recorded, including the identity of the investigator who made the modification, the date, the old and the new value.

Staff members were trained and certified before being involved actively in the study. Certification included ability to conduct interviews, to perform phlebotomy and to process blood samples, to accurately measure anthropometric and BP levels and to enter data into electronic databases. The accurateness of the data was checked by an external quality control organization (PRN, North Hampshire, United Kingdom).

Finally, the 'Laboratoire Central du CHUV' is ISO 9001 certified and is regularly checked by the "Centre Suisse de Contrôle de Qualité" (CSCQ – Swiss Centre for Quality Control).

Power estimates and type 1 error rate

Power calculations were done using the program Quanto v1.1 (James Gauderman, University of Southern California, USA) [15]. Power for unmatched case-control studies was estimated using an arbitrary allelic frequency ranging from 0.1 to 0.5, an additive model, a disease prevalence of 50% and a type 1 error rate of 10^-7, which was used as a correction for multiple testing taking into account 500'000 genetics markers. Curves were drawn for estimated genetic effect sizes (odds ratios) of 1.2 to 1.8. To estimate power in a continuous trait analysis, SBP was used as a continuous outcome for independent subjects, with an additive model and a type 1 error rate of 10^-7. Curves were drawn for various minor allele frequencies (0.1 to 0.5). To minimize false positive results, the most significant genetic markers will be replicated in independent samples.

Statistical analysis

Statistical analyses were performed using Stata 9.1 (Stata Corp, College Station, USA). Results were expressed as mean ± standard deviation (SD) or as number of subjects and (percentage). Data for age group 35–75 years and for the canton de Vaud were extracted from the MONICA population surveys and used to assess trends [16]. Age at sampling was used for comparisons between the initial population, the random sample and the CoLaus study population, whereas age at examination was used to describe the CoLaus Study group characteristics. Comparisons were performed using Student's t-test or chi-square test for quantitative and categorical variables, respectively. Statistical significance was assessed for p < 0.05.

Sponsoring

The Study was sponsored in part by GlaxoSmithKline and all participants were duly informed about this sponsorship and consented for the use of biological samples and data by GlaxoSmithKline and its subsidiaries.

Recruitment

A total of 8121 individuals volunteered to participate in the study. Among these subjects, the first 6,738 were invited to attend the clinic and completed the examination. 549 participants (8.1%) were not of Caucasian ethnicity and were excluded from the CoLaus study and 6,189 participants met the inclusion criteria (including ethnicity) and were included in the CoLaus study. As the number of subjects who agreed to participate (8,121) was higher than the number of subjects initially planned for the CoLaus study (6,000), 1,383 could not be included into the study although they were willing to participate. One subject withdrew after consent due to personal reasons. Overall, the final CoLaus sample (n = 6,188) represents 43% of the eligible responders, 41% of all the responders and 31% of the initially sampled population.

Socio-economic characteristics of the CoLaus study participants

Clinical characteristics of the CoLaus Study participants

CVRFs within the CoLaus study

Smoking was reporter by 27% of the participants; the prevalence of current smoking was higher in men and tended to decrease with age, from 35.3% among 35–44 year olds to 20.7% among 65–75 year olds in men (p < 0.001), the corresponding figures being 28.1% and 14.7% in women (p < 0.001). Interestingly, in the 45–54 age class, prevalence of smoking was higher in women than in men (30.7% vs. 28.8% (p < 0.001)).

The prevalence of hypertension was 36.7% overall, was higher in men and increased with age: 18.3% and 75.1% in men aged 35–44 and 65–75, respectively (p < 0.001); the corresponding numbers in women were 9.9% and 59.1% (p < 0.001). Among hypertensive subjects, 50.1% were currently taking anti-hypertensive medication. Treatment for hypertension was more frequent in female subjects and increased with age (from 32.2% to 59.4%). Of the treated hypertensive subjects more than half (52.0%) had BP levels ≥ 140/90 mmHg.

High LDL cholesterol, high triglyceride and low HDL cholesterol levels were seen in 20.8%, 12.5% and 2.8% of the subjects, respectively and these conditions were also more prevalent in men. The prevalence of high HDL cholesterol was 53.4% and higher in women than men (71.6% vs. 33.2%). Overall, one third of the sample had dyslipidemia, the prevalence of which was higher in men (42.7%) than in women (26.6%).

Use of statin therapy in subjects with high LDL cholesterol was 5.6%. Prevalence of statin therapy was 69% in subjects after myocardial infarction, 33% in subjects after stroke and 70% in subjects after coronary artery bypass graft. Target level of LDL-cholesterol for secondary prevention, as recommended (< 2.6 mmol/l), was achieved in 28%, 38%, 28% and 33% of the subjects with diabetes, myocardial infarction, stroke and coronary artery bypass graft, respectively.

The overall prevalence of diabetes was 6.6%, and was higher in men. Nine subjects (2.2%) reported to have T1DM. The prevalence of diabetes increased with age, from 2.5% to 17.2% in men aged 35–44 and 65–75, respectively; the corresponding numbers were 1.2% and 9.0% in women with a peak prevalence of 17.2% in men aged > 65. Roughly a third were newly diagnosed diabetics (31.5% for women and 34.7% for men, respectively). The prevalence of IFG in the CoLaus population was 9.8% and was higher in men than in women (14.3% vs. 5.6%).

Treatment for diabetes was present in 73.9% of diabetic subjects but nearly all known diabetics were treated (96.3%). Of the treated diabetic subjects 63.9% had a fasting blood glucose ≥ 7 mmol/L, at the time of their visit.

Finally, 0.7% of women and 2.6% of men reported a personal history of myocardial infarction (p < 0.001); conversely, no gender differences were found regarding personal history of stroke: 1.0% of women vs. 1.3% of men, p = 0.16.

Power estimates

The power of the CoLaus study for an unmatched case-control design to study genotype/phenotype associations depends on the number of cases/controls and estimated effect size [Figure 2]. For hypertension with 2268 cases, the estimated power is 0.9 for a 1.4 effect size and 0.5 for a 1.3 effect size. Cases for main CVRFs in the CoLaus study are: dyslipidemia: 2021, obesity: 974, smoking (> 25 cigarettes/day): 746, Type 2 diabetes: 398, coronary heart disease: 262, low HDL: 170. For continuous trait analysis, the example of SBP was taken [Figure 3]. For allelic frequencies of 0.2 to 0.4, the study has an estimated power of >0.8 to detect BP variations of 2.0 – 2.3 mm Hg.

Perspectives

A more comprehensive characterization of the Colaus participants is currently ongoing. First all participants aged 35 to 65 were solicited to undergo a psychiatric investigation based on a semi-structured diagnostic interview. In addition, 500 subjects were assessed for CV functional measurements. Indeed several previous studies have revealed associations between mood disorders and, in particular depression, and CVRFs and CV diseases (reviewed in [36]). The availability of a simultaneous CV and psychiatric phenotype will allow us to further explore the epidemiologic and potentially a genetic basis for this association. Finally, 500 randomly selected non-diabetic individuals underwent a 2 hour glucose tolerance test. The results from these investigations will be reported separately.

A longitudinal follow-up of all participants in the CoLaus study is planned and shall provide essential data for trends over time of major CVRFs and on incident cases of CV diseases. In the current cross-sectional study, participants were asked whether they consented to be contacted for follow-up, with > 90% favorable responses.