Dataset Information

Within-person variability in calculated risk factors: comparing the aetiological association of adiposity ratios with risk of coronary heart disease.

ABSTRACT:

Background

Within-person variability in measured values of a risk factor can bias its association with disease. We investigated the extent of regression dilution bias in calculated variables and its implications for comparing the aetiological associations of risk factors.

Methods

Using a numerical illustration and repeats from 42,300 individuals (12 cohorts), we estimated regression dilution ratios (RDRs) in calculated risk factors [body-mass index (BMI), waist-to-hip ratio (WHR), and waist-to-height ratio (WHtR)] and in their components (height, weight, waist circumference, and hip circumference), assuming the long-term average exposure to be of interest. Error-corrected hazard ratios (HRs) for risk of coronary heart disease (CHD) were compared across adiposity measures per standard-deviation (SD) change in: (i) baseline and (ii) error-corrected levels.

Results

RDRs in calculated risk factors depend strongly on the RDRs, correlation, and comparative distributions of the components of these risk factors. For measures of adiposity, the RDR was lower for WHR [RDR: 0.72 (95% confidence interval 0.65-0.80)] than for either of its components [waist circumference: 0.87 (0.85-0.90); hip circumference: 0.90 (0.86-0.93) or for BMI: 0.96 (0.93-0.98) and WHtR: 0.87 (0.85-0.90)], predominantly because of the stronger correlation and more similar distributions observed between waist circumference and hip circumference than between height and weight or between waist circumference and height. Error-corrected HRs for BMI, waist circumference, WHR, and WHtR, were respectively 1.24, 1.30, 1.44, and 1.32 per SD change in baseline levels of these variables, and 1.24, 1.27, 1.35, and 1.30 per SD change in error-corrected levels.

Conclusions

The extent of within-person variability relative to between-person variability in calculated risk factors can be considerably larger (or smaller) than in its components. Aetiological associations of risk factors should be compared through the use of error-corrected HRs per SD change in error-corrected levels of these risk factors.

SUBMITTER: Wormser D

PROVIDER: S-EPMC3733701 | biostudies-literature | 2013 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Within-person variability in calculated risk factors: comparing the aetiological association of adiposity ratios with risk of coronary heart disease.

Wormser David D White Ian R IR Thompson Simon G SG Wood Angela M AM

International journal of epidemiology 20130601 3

<h4>Background</h4>Within-person variability in measured values of a risk factor can bias its association with disease. We investigated the extent of regression dilution bias in calculated variables and its implications for comparing the aetiological associations of risk factors.<h4>Methods</h4>Using a numerical illustration and repeats from 42,300 individuals (12 cohorts), we estimated regression dilution ratios (RDRs) in calculated risk factors [body-mass index (BMI), waist-to-hip ratio (WHR), ...[more]

PMID: 23918853

Similar Datasets

Project description:Our objective was to quantify short-term total within-person variability in standard and nontraditional kidney measures using national data.Repeated examination study of serum and urine kidney measures.Participants 18 years or older in the Third National Health and Nutrition Examination Survey (NHANES III) who had repeated blood and urine samples collected during visits occurring approximately 18 days apart.Standardized serum creatinine, standardized cystatin C, ?-trace protein (BTP), ?(2)-microglobulin (B2M), and urine albumin and creatinine. We calculated the within-person coefficient of variation (CV(w)), which includes both biological and analytical variability. We also evaluated the impact of variability on estimates of the prevalence of reduced estimated glomerular filtration rate and albuminuria.Serum cystatin C level demonstrated the lowest short-term within-person variability (CV(w) = 6.8%). Serum creatinine and B2M levels (CV(w) = 7.6% and 8.4%, respectively) also had low variability. BTP level had the most variability of the serum markers (CV(w) = 11.6%). As expected, urine albumin and urine creatinine measurements showed high variability (CV(w) >30% for both); however, albumin-creatinine ratio performed much better than either measure alone, with CV(w) of 11.3%. The effect of short-term variability on the prevalence of reduced estimated glomerular filtration rate was moderate, with an ~20% lower prevalence when defined based on single measurements compared to repeated application of the same test approximately 18 days apart. Repeated testing for albuminuria had a larger effect, showing a 33% lower prevalence of albuminuria when repeated testing was applied.Only 2 measurements available. General population with low prevalence of kidney disease.Our results suggest that creatinine, cystatin C, and B2M levels have similarly low short-term variability. BTP level was more variable compared with the other serum filtration markers. Urine albumin and creatinine levels were highly variable and may benefit from repeated assessments to reduce the misclassification of albuminuria.

Project description:Background. In recent years, researchers have investigated the relationship between facial width-to-height ratio (FWHR) and a variety of threat and dominance behaviours. The majority of methods involved measuring FWHR from 2D photographs of faces. However, individuals can vary dramatically in their appearance across images, which poses an obvious problem for reliable FWHR measurement. Methods. I compared the effect sizes due to the differences between images taken with unconstrained camera parameters (Studies 1 and 2) or varied facial expressions (Study 3) to the effect size due to identity, i.e., the differences between people. In Study 1, images of Hollywood actors were collected from film screenshots, providing the least amount of experimental control. In Study 2, controlled photographs, which only varied in focal length and distance to camera, were analysed. In Study 3, images of different facial expressions, taken in controlled conditions, were measured. Results. Analyses revealed that simply varying the focal length and distance between the camera and face had a relatively small effect on FWHR, and therefore may prove less of a problem if uncontrolled in study designs. In contrast, when all camera parameters (including the camera itself) are allowed to vary, the effect size due to identity was greater than the effect of image selection, but the ranking of the identities was significantly altered by the particular image used. Finally, I found significant changes to FWHR when people posed with four of seven emotional expressions in comparison with neutral, and the effect size due to expression was larger than differences due to identity. Discussion. The results of these three studies demonstrate that even when head pose is limited to forward facing, changes to the camera parameters and a person's facial expression have sizable effects on FWHR measurement. Therefore, analysing images that fail to constrain some of these variables can lead to noisy and unreliable results, but also relationships caused by previously unconsidered confounds.

Project description:BackgroundVitamin D nutrition research requires accurate measures of circulating 25-hydroxyvitamin D. Our objectives were to test whether a diurnal fluctuation in blood-spot concentrations of 25-hydroxyvitamin D can be demonstrated statistically in a single individual, and whether such fluctuation is affected by the pre-dose versus post-dose timing of the blood draw.Case presentationThe participant in this case study was a generally healthy Caucasian woman in her 40s who has taken 5000 IU vitamin D3 supplement at midday for over 1 year. Each blood sample was drawn individually from a finger prick onto filter paper at morning, midday, or night, on 4 days (three groups of five individual blood samples per collection day). On days 1 and 2, the midday samples were collected approximately 1 hour after the supplement was taken; on days 3 and 4, the midday samples were collected within an hour prior to supplementation (the classical, daily "trough" value for a drug). There was a significant daily pattern of variation in 25-hydroxyvitamin D concentrations (analysis of variance p ≤ 0.02 for 3 of the 4 days): peak midday mean 25-hydroxyvitamin D was approximately 20% higher than in the morning, and approximately 13% higher than in the evening. Trough sampling produced no significant difference in 25-hydroxyvitamin D compared to sampling an hour after the dose. An incidental finding was that acute illness during the study was related to acutely lower 25-hydroxyvitamin D at every sampling time in the day (p < 0.00001).ConclusionsThere was a consistent diurnal variation in 25-hydroxyvitamin D, with the peak at midday. There was no difference between trough versus post-dose blood draws. Acute illness may acutely lower serum 25-hydroxyvitamin D levels. Because within-person, within-day variability in 25-hydroxyvitamin D is approximately 20%, sampling time introduces systematic error in vitamin D nutritional assessment that is bigger than random analytical error or choice of assay method.

Dataset Information

Within-person variability in calculated risk factors: comparing the aetiological association of adiposity ratios with risk of coronary heart disease.

Background

Methods

Results

Conclusions

Publications

Within-person variability in calculated risk factors: comparing the aetiological association of adiposity ratios with risk of coronary heart disease.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets