Cohort Studies

Large-scale cross-sectional and cohort studies have transformed our understanding of the genetic and environmental determinants of health outcomes. However, the representativeness of these samples may be limited-either through selection into studies, or by attrition from studies over time. Here we explore the potential impact of this selection bias on results obtained from these studies, from the perspective that this amounts to conditioning on a collider (i.e. a form of collider bias). Whereas it is acknowledged that selection bias will have a strong effect on representativeness and prevalence estimates, it is often assumed that it should not have a strong impact on estimates of associations. We argue that because selection can induce collider bias (which occurs when two variables independently influence a third variable, and that third variable is conditioned upon), selection can lead to substantially biased estimates of associations. In particular, selection related to phenotypes can bias associations with genetic variants associated with those phenotypes. In simulations, we show that even modest influences on selection into, or attrition from, a study can generate biased and potentially misleading estimates of both phenotypic and genotypic associations. Our results highlight the value of knowing which population your study sample is representative of. If the factors influencing selection and attrition are known, they can be adjusted for. For example, having DNA available on most participants in a birth cohort study offers the possibility of investigating the extent to which polygenic scores predict subsequent participation, which in turn would enable sensitivity analyses of the extent to which bias might distort estimates.

The German National Cohort (NAKO) is a multidisciplinary, population-based prospective cohort study that aims to investigate the causes of widespread diseases, identify risk factors and improve early detection and prevention of disease. Specifically, NAKO is designed to identify novel and better characterize established risk and protection factors for the development of cardiovascular diseases, cancer, diabetes, neurodegenerative and psychiatric diseases, musculoskeletal diseases, respiratory and infectious diseases in a random sample of the general population. Between 2014 and 2019, a total of 205,415 men and women aged 19-74 years were recruited and examined in 18 study centres in Germany. The baseline assessment included a face-to-face interview, self-administered questionnaires and a wide range of biomedical examinations. Biomaterials were collected from all participants including serum, EDTA plasma, buffy coats, RNA and erythrocytes, urine, saliva, nasal swabs and stool. In 56,971 participants, an intensified examination programme was implemented. Whole-body 3T magnetic resonance imaging was performed in 30,861 participants on dedicated scanners. NAKO collects follow-up information on incident diseases through a combination of active follow-up using self-report via written questionnaires at 2-3 year intervals and passive follow-up via record linkages. All study participants are invited for re-examinations at the study centres in 4-5 year intervals. Thereby, longitudinal information on changes in risk factor profiles and in vascular, cardiac, metabolic, neurocognitive, pulmonary and sensory function is collected. NAKO is a major resource for population-based epidemiology to identify new and tailored strategies for early detection, prediction, prevention and treatment of major diseases for the next 30 years.

ObjectivesMotivated by recent calls to use electronic health records for research, we reviewed the application and development of methods for addressing the bias from unmeasured confounding in longitudinal data.Study design and settingMethodological review of existing literature. We searched MEDLINE and EMBASE for articles addressing the threat to causal inference from unmeasured confounding in nonrandomized longitudinal health data through quasi-experimental analysis.ResultsAmong the 121 studies included for review, 84 used instrumental variable analysis (IVA), of which 36 used lagged or historical instruments. Difference-in-differences (DiD) and fixed effects (FE) models were found in 29 studies. Five of these combined IVA with DiD or FE to try to mitigate for time-dependent confounding. Other less frequently used methods included prior event rate ratio adjustment, regression discontinuity nested within pre-post studies, propensity score calibration, perturbation analysis, and negative control outcomes.ConclusionWell-established econometric methods such as DiD and IVA are commonly used to address unmeasured confounding in nonrandomized longitudinal studies, but researchers often fail to take full advantage of available longitudinal information. A range of promising new methods have been developed, but further studies are needed to understand their relative performance in different contexts before they can be recommended for widespread use.

The CORDELIA Study (Collaborative Cohorts Reassembled Data to Study Mechanisms and Long-term Incidence of Chronic Diseases) combines 35 Spanish population cohorts to investigate the clinical, environmental, genetic, and omics determinants of cardiovascular disease in the Southern European population. It aims to conduct the largest genome-wide association study to date on cardiovascular disease in this population, improve predictions of cardiovascular incidence using genomic and clinical data, and identify subgroups that would benefit most from targeted pharmacological and lifestyle interventions. CORDELIA includes 196,632 individuals (ages 18-84, 54% female, 96% born in Spain, 20% with higher education, recruited from 1989 to 2020, with follow-up periods ranging from 5 to 30 years), with DNA samples available for 117,342 participants (60%). Of the participants, 24% were current smokers, 43% hypertensive, 11% diabetic, 15% medicated with lipid-lowering drugs, 44% overweight, and 27% obese. If not already available, genotyping is being performed using the Axiom™ Spain Biobank array (~ 750,000 variants, including 115,000 specific and 50,000 rare functional variants from the Spanish population). The cohort also includes incident events (coronary heart disease, stroke, heart failure, peripheral artery disease, hypertension, diabetes); date and cause of death; and harmonized data on risk factors (body mass index, waist circumference, lipid profile, blood pressure, glucose, creatinine), lifestyle (smoking, physical activity, diet, alcohol), and socioeconomic status. 99,019 participants (50%) also provide plasma samples. CORDELIA will significantly contribute to understanding the complex interplay of risk factors contributing to cardiovascular disease and advance the fields of precision medicine and public health in Southern European individuals.