The Rotterdam Study: objectives and design update

Research on the epidemiology of cardiovascular diseases focuses on three primary areas of interest: studies on risk factors for atherosclerosis and coronary heart disease, studies on the detection of subjects at high risk of coronary heart disease, and studies on cardiovascular conditions in older age.

Two groups of putative risk factors for atherosclerosis and coronary heart disease are included. The first are endocrine factors, including estrogens and androgens, insulin and insulin-like growth hormone I, and thyroid gland and adrenal gland hormones. The second group contains factors involved in haemostasis, inflammation and endothelial function. Research also focuses on genetic factors in these areas in relation to risk of coronary heart disease.

The ability of classical cardiovascular risk factors to identify subjects at high risk of coronary heart disease is limited. Risk stratification may be improved when based on the presence of atherosclerosis. To this end, repeated measurements of non-coronary atherosclerosis and measurements of coronary, carotid and aortic arch calcification have been included in the study.

Another line of research focuses on cardiovascular diseases in the elderly that are in large part the consequence of ischemic heart disease, like heart failure and atrial fibrillation. An important topic in this area is the early diagnosis of heart failure using echocardiographic assessment of asymptomatic systolic and diastolic dysfunction of the left ventricle. Atrial fibrillation is another major chronic condition frequent at older age. Examination of the determinants and prognosis of atrial fibrillation is part of this research line.

Findings on determinants of atherosclerosis and coronary heart disease include the relation between hypothyroidism and myocardial infarction [4] and the association of genetic variation in the estrogen receptor alpha gene with risk of myocardial infarction in women [5]. The study demonstrated that high levels of CRP [6] and lipoprotein-associated phospholipase A2 activity [7] were associated with risk of coronary heart disease and stroke. Genetic findings in this area include a positive association of coronary heart disease with genetic variation in the complement factor H gene [8] but not with variation in the CRP gene [6].

The study showed that carotid intima-media thickness and plaques measured by ultrasound, the ankle-arm index, and aortic calcification detected by X-ray are independent predictors of risk of coronary heart disease [9]. The study also showed that vessel wall stiffness, as measured by aortic pulse wave velocity (PWV), improved prediction of cardiovascular disease beyond the classical risk factors [10]. The Rotterdam Study was the first to publish results from population-based research on the predictive value for coronary heart disease of coronary calcification assessed by electron-beam CT [11].

The study enabled accurate assessment of the incidence and lifetime risk of heart failure and atrial fibrillation in an elderly population [12, 13]. It was shown that atherosclerosis and inflammation are also involved in these conditions [14, 15].

Repeated measures of non-coronary atherosclerosis included carotid intima-media thickness and plaques by ultrasound, the ankle-arm index and aortic calcification by X-ray [9]. Electron-beam CT and multi-detector CT were used to accurately quantify calcification in the coronary, aortic arch and carotid arteries [11]. In the second additional cohort of the Rotterdam Study (RS Young1, see Fig. 1), the measurement of plaque vulnerability with high-resolution MRI of the carotid arteries will be added. Other outcome measures include electrocardiography, echocardiography and measures of vessel wall stiffness (PWV and carotid distensibility) [10].

Determinants were assessed by physical examinations, collection of blood samples, and by questionnaires and interview. The role of genetic factors is studied using the candidate gene approach and more recently the genome wide association study.

Clinical cardiovascular outcomes are collected during our continuous follow-up and include non-fatal myocardial infarction and cardiac death, revascularizations, heart failure and atrial fibrillation. The methods of diagnosis of these outcomes have been described in previous papers [6, 12, 13].

Neuroepidemiologic research in the Rotterdam Study focuses on the frequency, etiology and early recognition of the most frequent neurologic diseases in the elderly, including dementia (in particular Alzheimer disease), Parkinson disease and stroke. In neurodegenerative and cerebrovascular disorders clinical symptoms typically become manifest late in the disease course, the occurrence of clinical disease does not reflect the underlying spectrum of disease-related pathology, and most of the clinical syndromes are etiologically heterogeneous. Therefore, an additional research focus is on the causes and consequences of pre-symptomatic brain pathology that can be assessed with non-invasive imaging modalities.

Neurodegenerative and cerebrovascular diseases are highly frequent in the elderly. The prevalence increases from age 55 to 65 years to age 90 years and above from less than 1% to over 40% for dementia [16], from less than 0.5% to more than 4% for Parkinson disease [17], and from approximately 1% to nearly 10% for stroke. The incidence figures follow this pattern of a strong increase with age over the entire age range, with the age-specific incidence of dementia being identical for men and women at least until the age of 85 [18] but with men having a higher age-specific incidence of both stroke and Parkinson disease than women throughout the age range [19, 20].

Vascular pathology and vascular risk factors are associated with worse cognitive performance [21], which also translates in people with vascular pathology or risk factors for vascular disease having an increased risk of dementia, including Alzheimer disease [22]. Moreover, several life style factors are associated with the risk of dementia and Alzheimer disease [23‐25], suggesting that onset of dementia may at least partly be delayed or prevented. Commonly used drugs may have a role in this [26].

The classical risk factors for stroke also predict risk of stroke in the Rotterdam Study [27]. More recently identified risk factors, including inflammatory markers, may be etiologically relevant but thus far add little to the identification of people at risk [28]. Possibly underlying this is that a large amount of stroke goes clinically undetected [29]. Nearly 20% of elderly people have at least one silent brain infarct, and thereby a nearly fourfold increased risk of clinical stroke, a more than doubled risk of dementia including Alzheimer disease, and an increased risk of depression [29].

Neuroimaging reveals that brain pathology is widespread [30] and can go clinically undetected for a long time. In addition to the silent infarcts, many apparently healthy elderly have ischemic changes in their cerebral white matter that are associated with an increased risk of dementia, stroke and depression [31]. Also brain atrophy, especially of the hippocampus, is already present years before onset of even the earliest sign of cognitive impairment or subjective complaints [32]. This emphasizes the need to shift the attention in etiologic research of neurodegenerative and cerebrovascular disease to the causes of pre-symptomatic and underlying brain changes.

The aim of the psychiatric research in the Rotterdam Study is to investigate the determinants, correlates and consequences of common psychiatric problems in the elderly. The focus has been on depressive disorders but anxiety disorders, sleep disturbances and complicated grief are also being studied.

The one-month prevalence of both major and minor depression at baseline was 1.5%, the prevalence of anxiety disorder was 8.2%.

We found evidence for the vascular depression hypothesis. More severe coronary and extra-coronary atherosclerosis were associated with a higher prevalence of depression, as were cerebral haemodynamic changes [34, 35]. However, we could not rule out that earlier depressive episodes may have contributed to the development of atherosclerosis. Moreover, our data did not support a specific symptom profile of vascular depression as previously defined [36].

Most of our studies of nutritional and metabolic factors provided no support for an etiological role in depression. Folate deficiency, homocysteine, fatty acid ratio (omega-6/omega-3) and vitamin E had little or no effect on the risk of depression after control for health and health related behaviour [37‐39]. In contrast, low levels of vitamin B12 and high levels of inflammatory protein IL-6 convey a strong risk for depression, although the latter effect may be confined to high-risk groups [40].

Several SNPs have been investigated, mostly with negative results. However, we found that an ER-alpha polymorphism had a substantial effect on anxiety but not on depression in women [41].

In the first years of the Rotterdam Study psychiatric data collection was very limited. However, from the third examination onwards, which began in 1997 (see Fig. 1), depressive symptoms and disorders are have been ascertained in all participants. The 4,603 participants with valid depression scores in this examination thus constitute the baseline population for longitudinal studies of depression. Because assessment of anxiety disorders, sleeping disturbances and complicated grief were added in the fourth examination, they can currently be studied only cross-sectionally.

The primary objectives of the ophthalmological part of the Rotterdam Study are to study frequency and risk factors of common and invalidating eye diseases with emphasis on age-related macular degeneration (AMD) and primary open angle glaucoma (POAG). Another line of research focuses on the associations between retinal vessel abnormalities and cardiovascular and neurologic diseases.

AMD was the main cause of blindness in the high age-group [47]. Major risk factors that were found for AMD are smoking, atherosclerosis, hyperopia [48‐50]. The APO-E ε-4 allele showed an inverse association with AMD [51]. First-degree relatives of patients with late ARM developed ARM at an increased rate at a relatively young age [52]. The heterogeneity of genetic risk among AMD families is considerable, and the proportion of high-risk families is relatively small [53]. The anticipated protective effect of cholesterol-lowering drugs on AMD could not be substantiated [54]. A protective effect on AMD was demonstrated in participants using a diet rich of antioxidants and zinc intake in our study population. Support for a multifactorial origin of AMD was found in the cumulative effects of aberrant CFH and CRP genes, inflammations and smoking as genetic and environmental stimulators of the complement cascade [55].

The prevalence estimates of POAG in Rotterdam was highly dependent on the applied criteria [56]. That was a major reason for proposing a classification system without final, subjective adjudication [57]. Systemic blood pressure and hypertension were associated with elevated intraocular pressure but not with prevalent POAG [58]. Relatives of patients with POAG have a strongly increased risk of glaucoma. Enlarged cup-disc ratio was the earliest and most prominent feature of familial aggregation [59]. The incidence of OAG rises significantly with age. Most of the patients with incident OAG were unaware of having OAG [60]. The incidence of visual field loss rises fivefold between 55 and 80 years to 20/1,000 person years in the general population [61]. Glaucoma is the main cause for visual field loss, followed by stroke [62].

Larger retinal venular diameters are associated with generalized atherosclerosis, inflammation and cholesterol levels. Retinal venular diameters are variable and may play their own independent role in predicting cardiovascular disorders [62]. Dilated retinal venules at baseline were predictive for stroke, cerebral infarction, dementia, white brain matter lesions, impaired glucose tolerance, diabetes mellitus and mortality [63, 64]. In systemic hypertension, the arteriolar diameters are narrow and arteriolar and venular narrowing precedes the onset of hypertension [65].

Repeated ophthalmic measurements include best corrected ETDRS visual acuity, refractive error, Goldmann applanation tonometry, keratometry, slitlamp examination of the anterior segment and visual field testing. In pharmacological mydriasis we make colour photographs of the lens, 35 degree photography of the macular area, and simultaneous stereoscopic imaging of the optic disc and macular area. Digital photography of the macular area and optic disc are added since the fourth follow-up examination. Scanning laser ophthalmoscopy (HRT II) measurements of the optic disc, macular pigment density measurements and optical coherence tomography of the macular area and optic disc were added during follow-up.

For the assessment of AMD we use stereoscopic 35° macular images centred on the fovea. The images are graded according to the International Classification and Grading System for AMD [66].

The procedure for assessing POAG includes Goldmann applanation tonometry, visual field screening, ophthalmoscopy and stereoscopic fundus photography [57].

Assessment of retinal vascular diameters is done by analyses of digitized fundus transparencies with a semi automated system [58, 67].

The first objective of the laboratory team is to collect, store and manage the biological tissues sampled in the Rotterdam Study. The second objective of the group concerns genotyping and assessment of biomarkers.

Both from biomarker studies as well as molecular genetic studies important findings have emerged. For example, among the biomarker analyses the study documenting the relationship between homocysteine and osteoporosis has been novel [68] and since widely replicated. Several candidate gene studies have also yielded major new insights coming from both exploratory studies as well as from collaborative replication efforts across all disease-oriented research lines in the Rotterdam Study. A unique feature of the Rotterdam Study is exploited by studying the relationship between pleiotropic gene variants and not one but multiple diseases and disease-related endpoints. For example, the studies on the promoter region of the IGF-1 gene revealed a series of consistent associations ranging from birth weight to diabetes [69], while other consistent associations involve the estrogen receptor alpha (ESR1) gene in relation to osteoporosis [70], osteoarthritis, height, myocardial infarction [5], age-at-menopause, and depression.

The Rotterdam Study is involved as a major collaborative centre for studies for novel genes coming from genome-wide analysis (GWA) studies, including CFH in age-related macula degeneration [71], NOS1AP in QT interval [72], and several SNPs involved in height, type 2 diabetes, and breast cancer. Finally, Rotterdam Study investigators play a leading role in the emerging large global consortia focussing on the contribution of complex disease gene variants by prospective meta-analyses across many epidemiological cohorts, such as for osteoporosis in the GENOMOS Study and EUROSPAN.

A major objective of the pharmaco-epidemiologic studies is to investigate the role of drugs as determinants of disease in the Rotterdam Study. This includes studying efficacy and effectiveness of drugs, as well as adverse reactions to drugs.

Important findings have been published on pharmaco-epidemiological topics concerning the main outcomes in the Rotterdam Study. Studies about the association between dementia and antihypertensive drugs [75] and NSAIDs [26] have strongly suggested a protective effect of both groups of drugs. Several studies have been performed on cardiovascular topics [76‐78]. In one of these studies, NSAIDs were associated with an increased risk of heart failure [79]. In line with the suspicion that QTc-prolonging drugs may cause sudden cardiac death, it was demonstrated in the Rotterdam Study that a prolonged QTc is indeed an important risk factor [77]. Furthermore, in one study it was demonstrated that high-dose corticosteroids increase the risk of atrial fibrillation [78]. In the important area of locomotor diseases, studies have demonstrated that thiazide diuretics protect against hip fracture [79] and that statins reduce the risk of vertebral fracture [80]. On the other hand, the risk that long-term use of certain NSAIDs may aggravate signs of osteoarthritis has been emphasized [81]. In the area of ophthalmological diseases, a protective effect of cholesterol-lowering agents on macular degeneration has been suggested [82]. In other areas, such as pharmacogenetics and other causes of interactions between drugs, several important findings have been published [83‐91].

For several reasons, a drug is a highly attractive determinant in clinical epidemiology. First, drugs are probably the most important therapeutic intervention in health care. Despite rigorous clinical research before registration, many important effects of drugs are discovered after marketing. Second, all marketed drugs have proven biological activity, meaning that it concerns a determinant which really matters. Third, and as a consequence of the availability of complete medication histories in Dutch health care, the role of drug exposure can be assessed in a detailed way.

In the Rotterdam Study, there is an almost complete coverage of the population as of 1 January, 1991 thanks to the fact that all pharmacies which serve the Ommoord region are on one computer network. To date, almost 3 million prescriptions have been delivered to the population of the Rotterdam Study and of each prescription, details are available about the product name and contents, ATC-code, dosage and duration of drug therapy.

Drugs are a group of determinants, which can be studied in association with a large variety of diseases. In the Rotterdam Study, there is a strong interest in the association between drugs and the cardiovascular, neurological, endocrinological and ophthalmological diseases, which have been the main topics since starting. However, there is also important information about the association with psychiatric diseases, cancer, and chronic obstructive pulmonary disease. Moreover, important information about secondary outcomes, such as drug blood levels, other laboratory information, and information about hospital discharge diagnoses, is gathered on a continuous basis to facilitate pharmaco-epidemiological studies.