Epidemiological studies outcome measures

Several recent epidemiological studies have involved examination of populations that consume unusually high levels of fish. One of these, conducted in the islands of the Seychelles, has not so far revealed behavioral and learning impairments in children whose mothers exhibited mercury levels (measured in hair) higher than those typically seen in the United States and European countries. But another study, conducted in the Faroe Islands, turned up evidence of cognitive and behavioral impairments in children. Scientists have struggled to understand why two well-done studies have produced such different outcomes, and some possible reasons have been suggested. The EPA and public health officials have acted on the basis of the Faroe data, out of both caution and also because they seem to be supported by other, more limited data, and by experimental studies. The debate is not so much about whether methylmercury is a developmental toxicant, but rather over the dose required. 

Pharmacokinetics has played a crucial and somewhat unusual role in the assessment of health risks from methylmercury. Some of the epidemiology studies of this fish contaminant involved the measurement of mercury levels in the hair of pregnant women, and subsequent measurements of health outcomes in their offspring (Chapter 4). Various sets of pharmacokinetic data allowed estimation of the level of methylmercury intake through fish consumption (its only source) that gave rise to the measured levels in hair. In this way it was possible to identify the dose-response relationship in terms of intake, not hair level. Once the dose-response relationship was established in this way, the EPA was able to follow its usual procedure for establishing an RfD (which is 0.1 ag/(kg b.w. day)). 

Chlorpyrifos provides an example of the utility of human pharmacokinetic models to estimate daily dose from biomonitoring data for a rapidly cleared pesticide. The urinary metabolite trichloro-2-pyridinol (TCP) is used in the NHANES study to monitor population exposure to chlorpyrifos (CDC 2005). Several epidemiologic studies have linked chlorpyrifos exposure to adverse birth outcomes through associations between urinary and blood biomarkers and have demonstrated maternal exposure and physiologic measurements in the neonate (Berkowitz et al. 2003, 2004 Whyatt et al. 2004 Needham 2005). 

First, this chapter will describe a conceptual framework to illustrate the special challenges posed because exposures assessed for epidemiologic studies must be relevant to the health outcome under investigation. Secondly, some of the most commonly applied epidemiological study designs will be introduced, with special emphasis on exposnre assessment issnes associated with the design. Thirdly, some widely applied exposure assessment approaches will be introduced, ranging from qualitative classifications of exposure to quantitative exposure assessment of pesticide concentrations. The influence of measurement error on measures of association between exposure and disease, such as the slopes of exposure-response relationships and risk or odds ratios, will be briefly reviewed. Finally, exposure proxies used in case-control studies of chronic effects of pesticide exposure will be reviewed and the concepts introduced earlier will be applied. 

Major limitations in many of the in vitro studies to date have been (1) the use of aglycones, when there is no evidence that the aglycone is absorbed and presented to the tissues, and (2) the use of concentrations well above those observed in plasma. Few studies to date have attempted to measure anthocyanin concentrations in different tissues. Research with anthocyanins has been slowed due to the lack of pure standard compounds, particularly of the anthocyanins and the availability of isoto-pically labeled anthocyanins, labeled so that the label is stable at different pH. Understanding any potential relationships to disease prevention has been limited because of the lack of availability of any database on the food content of anthocyanins. These data are being acquired in the U.S., allowing for estimation of daily intakes of anthocyanins from food intake data and for studying relationships to disease outcome in epidemiology studies. 

Another difficulty arises from the use of cumulative exposure (the product of exposure duration x intensity) as a surrogate exposure metric in the available studies. Finkelstein (1995) noted that the use of cumulative exposure requires the assumption that duration and intensity are equally important in determining the effective dose. Finkelstein further noted that if exposure estimates are inaccurate or inconsistently measured (which can be the case for many retrospective epidemiology studies), a finding of a statistically significant association between cumulative exposure and a health outcome can mislead one in having confidence in an apparent exposure-response relationship that is principally influenced by duration of exposure and not by exposure intensity. 

Several epidemiological studies on the effects of PPOA in humans have been conducted on workers. These studies did not examine developmental outcomes. A retrospective cohort mortality study demonstrated a statistically significant association between prostate cancer mortality and employment duration in the chemical facility of a plant that manufactures PPOA. However, this result was not observed in an update to the study in which more specific exposure measures were used. 

There are differences in the outcomes of these epidemiology studies on low level chronic exposures to methylmercury in foods. Davidson et al. (1998) report no adverse developmental effects associated with prenatal and postnatal exposure to methylmercury in fish in a Seychelles Island cohort of children at age 66 months (n=708). The exposure levels are reflected in maternal hair levels of 6.8 ppm for the prenatal exposure (SD=4.5, n=711) and children s hair levels of 6.5 ppm (SD=3.3, n=708) for both the prenatal and subsequent postnatal exposure. The age-appropriate main outcome measures included (1) the... 

The appropriate selection of a method for reconstruction and estimation is extremely important, whether for an exposure assessment, risk assessment, or epidemiological study. Some of the factors affecting the selection of an exposure reconstruction method include the characteristics of the outcome of interest (such as carcinogenic latency and dose of concern) the effectiveness of analytical methods to identify and measure the substance of interest (such as ability to distinguish from similar substances or interferences caused by mixtures of substances) and the consideration of exposures by multiple routes, including inhalation and dermal exposures. 

Understanding how people are exposed to soil contaminants is fundamental to any study investigating a possible link with health. That said, most health and epidemiological studies fail to appreciate, or identify, if and how people are exposed. For example, many studies of communities around landfill sites simply use proximity to the site of interest as a proxy measure of exposure. Often such an approach is unavoidable, as there may be insufficient data to allow for a proper exposure assessment. While this approach has some merit, the lack of a proper exposure assessment will mean that the outcomes of the study (whether positive or negative) will be significantly weakened. 

As noted by the lOM, Although a wide range of possible exposures might be associated with adverse health outcomes in PGW veterans, data on these exposures are often not available when they are available, they are poorly documented. This lack of exposure information is at the core of the frustration in obtaining answers from epidemiological studies. Self-reports of exposure and estimates of individual exposure from unit level measurements will be subject to so much error that they are likely to yield inconclusive results and additional questions. ... 

Several epidemiological studies have linked a protective effect against cardiovascular diseases and regular consumption of either red wine [7, 8, 98] or polyphenol-rich sources such as green tea [14,15,99], cocoa, or chocolate [16,17]. Since many cardiovascular diseases are associated with an impaired endothelial function, the evaluation of the endothelial function is of great importance. In humans, the endothelial function is often assessed by the flow-mediated dilatation (FMD). For this purpose, reactive hyperemia is induced after a transient ischemia in the upper arm using an inflatable cuff like that used for the measurement of blood pressure, while the diameter of the brachial artery is monitored continuously at the elbow level by ultrasound. The cuff is inflated tmtil the occlusion of the brachial artery, which is maintained for 5 min before the blood flow is released. The reperfusion then induces an endothelium-dependent dilatation of the brachial artery due to the increased shear stress, the most important physiological activator of eNOS. Impaired endothelial function, assessed by FMD, has been reported as an independent predictor of cardiovascular outcome in subjects with cardiovascular risk factors or established cardiovascular diseases [18]. 

This chapter deals with epidemiological studies of human Pb exposure using biomarkers of such exposure. These studies deal with various parameters and correlates of human Pb exposure from demographic and environmental measurements. While much of the available data has been generated in the United States, illustrative international studies are also presented where relevant. Later chapters deal with epidemiological studies of dose—response relationships for Pb and an array of toxicity outcome measures addressing organ and tissue toxicides. 

In the ten years since the results of the first epidemiological studies of the effects of low levels of lead were published, enormous progress has been made in almost all aspects of the research in study design, in the statistical techniques employed, in the outcome measures assessed, and in the measures of exposure. As a result of this we know a great deal more about the effects of low levels of environmental lead on developing children than we did a decade ago. The fact that there is still a lot that we do not know should not blind us to the achievements and progress that has been made. 

---