The effects of errors in lipid measurement and assessment

Developing a synthetic control group using electronic health records: Application to a single-arm lifestyle intervention study

The electronic health records (EHR) infrastructure offers a tremendous resource for identifying controls who match the characteristics of study participants in a single-arm trial. The objectives are to (1) demonstrate the feasibility of curating a synthetic control group for an existing study cohort through EHR data extraction and (2) evaluate the effect of a lifestyle intervention on selected cardiovascular health metrics. A total of 711 university employees were recruited between 2008 and 2012 to participate in a health partner intervention to improve cardiovascular health and were followed for five years. Data of nearly 8000 eligible subjects were extracted from the EHR to create a synthetic control cohort during the same study period. To minimize confounding, crude comparison, exact matching, propensity score matching, and doubly robust estimation were used to compare the selected cardiovascular health metrics at 1 and 5 years of follow-up. Blood pressure and body mass index improved in the intervention group compared to the EHR synthetic controls. The findings of changes in lipid measurements were somewhat unexpected. When analyzing the subgroup without lipid-lowering medications, the intervention group exhibited better control of cholesterol levels over time than did our synthetic controls. Some measurements in the EHR system may be more robust for synthetic selection than others. EHR synthetic controls can provide an alternative to estimate intervention effects appropriately in single-arm studies for these measurements.

Nonlinear and time-dependent effects of sparsely measured continuous time-varying covariates in time-to-event analysis

Many flexible extensions of the Cox proportional hazards model incorporate time-dependent (TD) and/or nonlinear (NL) effects of time-invariant covariates. In contrast, little attention has been given to the assessment of such effects for continuous time-varying covariates (TVCs). We propose a flexible regression B-spline-based model for TD and NL effects of a TVC. To account for sparse TVC measurements, we added to this model the effect of time elapsed since last observation (TEL), which acts as an effect modifier. TD, NL, and TEL effects are estimated with the iterative alternative conditional estimation algorithm. Furthermore, a simulation extrapolation (SIMEX)-like procedure was adapted to correct the estimated effects for random measurement errors in the observed TVC values. In simulations, TD and NL estimates were unbiased if the TVC was measured with a high frequency. With sparse measurements, the strength of the effects was underestimated but the TEL estimate helped reduce the bias, whereas SIMEX helped further to correct for bias toward the null due to "white noise" measurement errors. We reassessed the effects of systolic blood pressure (SBP) and total cholesterol, measured at two-year intervals, on cardiovascular risks in women participating in the Framingham Heart Study. Accounting for TD effects of SBP, cholesterol and age, the NL effect of cholesterol, and the TEL effect of SBP improved substantially the model's fit to data. Flexible estimates yielded clinically important insights regarding the role of these risk factors. These results illustrate the advantages of flexible modeling of TVC effects.

The challenging use and interpretation of blood biomarkers of exposure related to lipophilic endocrine disrupting chemicals in environmental health studies

The use of exposure biomarkers has been growing during the last decades, being considered the 'gold-standard' approach for individual exposure assessment to environmental chemicals. However, lipophilic endocrine disrupting chemicals (LEDC) have specific physicochemical and biological properties implying particular analytical challenges and interpretative caveats. The epidemiological literature is therefore afflicted by methodological inconsistencies and results divergences, in part due to recognised sources of exposure measurement error and misinterpretation of results. The aim of the present review is to identify external and endogenous sources of variability and uncertainty associated with the LEDC blood biomarkers in epidemiological studies. The dynamic nature of blood and an overview of the known mechanisms of transport, storage and partition of LEDCs in the organism are first described. The external sources of variability and uncertainty introduced at pre-analytical and analytical level are subsequently presented. Subsequently, we present some specific cases where the dynamics of lipids and LEDCs may be substantially modified and thus, the interpretation of biomarkers can be particularly challenging. The environmental obesogens as source of biomarkers variability is also discussed in the light of the most recent findings. Finally, different modelling approaches (statistical and pharmacokinetic models) proposed to improve the use and interpretation of biomarkers are appraised.

CE: Triglycerides: Do They Matter?

: Since the introduction of HMG-CoA reductase inhibitors, also known as statins, as an adjunct to diet in the treatment of hyperlipidemia and the greater emphasis placed on reducing low-density lipoprotein (LDL) cholesterol levels in the prevention of atherosclerosis and cardiovascular disease (CVD), there has been less focus on the value of lowering serum triglyceride levels. Many patients are aware of their "good" and "bad" cholesterol levels, but they may not be aware of their triglyceride level or of the association between high triglycerides and the development of CVD. In recent years, however, in light of the increasing incidences of obesity, insulin resistance, and type 2 diabetes, lowering triglyceride levels has gained renewed interest. In addition to the focus on lowering LDL cholesterol levels in CVD prevention, clinicians need to be aware of the role of triglycerides-their contribution to CVD, and the causes and treatment of hypertriglyceridemia.

Standardization of total cholesterol measurement in population surveys--pre-analytic sources of variation and their effect on the prevalence of hypercholesterolaemia

BACKGROUND

The estimation of population distribution of total cholesterol, as well as other blood lipids requires population surveys. Comparability of the estimates over time or between populations requires a predefined, standardized measurement protocol. This paper will assess the effect of variation in the pre-analytic procedures, on the estimation of population distribution of total cholesterol and the prevalence of hypercholesterolaemia. Implications of variation to real survey results are demonstrated on data collected during the WHO MONICA Project.

DESIGN AND METHODS

The extent (%) of pre-analytic variations were determined by literature review. Simulations were then used to estimate the effect of these variations on the distribution of total cholesterol values. Three populations were selected as examples for different population distributions of total cholesterol levels and variations resulting from seasonal and postural changes, use of tourniquet, and serum versus plasma collection were simulated both individually and in combinations.

RESULTS

Depending on the population distribution of total cholesterol, differences in pre-analytic procedures can explain a difference of up to 1.12 mmol/l in the mean total cholesterol between populations, and a difference up to 41% in the prevalence of hypercholesterolaemia (> or =6.5 mmol/l).

CONCLUSIONS

Variation in results on measured total cholesterol levels due to differences in procedures during the pre-analytic stage can diminish substantially the reliability and comparability of measurements among surveys over time and between populations. The effect of pre-analytic variation in the population estimates of total cholesterol can be extensively reduced by standardization and training.

CDC/AHA Workshop on Markers of Inflammation and Cardiovascular Disease

There is great interest in moving beyond established risk factors to consider markers of inflammation for the prediction of initial cardiovascular disease events. Inflammatory markers such as leukocyte count, serum amyloid A, C-reactive protein, and vascular outcomes in individuals free of cardiovascular disease at baseline are the key markers that have been investigated in the population setting. A meta-analysis of 11 prospective studies in asymptomatic individuals compared people in the bottom third of the C-reactive protein distribution with those in the top tertile. The authors reported an odds ratio of 2.0 (95% CI 1.6 to 2.5) for coronary heart disease among people in the top tertile. These results are among the strongest assembled thus far to recommend incorporating newer biomarkers into coronary heart disease risk estimation algorithms. A variety of issues should be considered and conditions satisfied before vascular disease risk factors are adopted into regular use. The type of vascular event and the follow-up interval are important features because results for short-, intermediate-, and long-term intervals may yield different results. The factors under consideration should have been standardized, and characteristics such as the variability of the measurements, correlations with established factors, evidence from observational studies and clinical trials, type of effect (linear, nonlinear, dichotomous), improvement in overall prediction (discrimination), generalization of results (calibration), and cost can affect the utility. Each of these issues needs to be considered and the effects on relative, absolute, and population-attributable risks described. In particular, we need to (1) develop sound strategies for implementing new testing and (2) demonstrate the benefit of testing by using the current foundation of prior probabilities developed from already-published risk factor assessments based on large population studies.