Jack C. Suggs

Frequency of mouthing behavior in young children

Young children may be more likely than adults to be exposed to pesticides following a residential application as a result of hand- and object-to-mouth contacts in contaminated areas. However, relatively few studies have specifically evaluated mouthing behavior in children less than 5 years of age. Previously unpublished data collected by the Fred Hutchinson Cancer Research Center (FHCRC) were analyzed to assess the mouthing behavior of 72 children (37 males/35 females). Total mouthing behavior data included the daily frequency of both mouth and tongue contacts with hands, other body parts, surfaces, natural objects, and toys. Eating events were excluded. Children ranged in age from 11 to 60 months. Observations for more than 1 day were available for 78% of the children. The total data set was disaggregated by gender into five age groups (10–20, 20–30, 30–40, 40–50, 50–60 months). Statistical analyses of the data were then undertaken to determine if significant differences existed among the age/gender subgroups in the sample. A mixed effects linear model was used to test the associations among age, gender, and mouthing frequencies. Subjects were treated as random and independent, and intrasubject variability was accounted for with an autocorrelation function. Results indicated that there was no association between mouthing frequency and gender. However, a clear relationship was observed between mouthing frequency and age. Using a tree analysis, two distinct groups could be identified: children ≤24 and children >24 months of age. Children ≤24 months exhibited the highest frequency of mouthing behavior with 81±7 events/h (mean±SE) (n=28 subjects, 69 observations). Children >24 months exhibited the lowest frequency of mouthing behavior with 42±4 events/h (n=44 subjects, 117 observations). These results suggest that children are less likely to place objects into their mouths as they age. These changes in mouthing behavior as a child ages should be accounted for when assessing aggregate exposure to pesticides in the residential environment.

The 1998 Baltimore Particulate Matter Epidemiology-Exposure Study: Part 1. Comparison of Ambient, Residential Outdoor, Indoor and Apartment Particulate Matter Monitoring

A combined epidemiological–exposure panel study was conducted during the summer of 1998 in Baltimore, Maryland. The objectives of the exposure analysis component of the 28-day study were to investigate the statistical relationships between particulate matter (PM) and related co-pollutants from numerous spatial boundaries associated with an elderly population, provide daily mass concentrations needed for the epidemiological assessment, and perform an extensive personal exposure assessment. Repeated 24-h integrated PM2.5 (n=394) and PM10 (n=170) data collections corresponding to stationary residential central indoor, individual apartment, residential outdoor and ambient monitoring were obtained using the same sampling methodology. An additional 325 PM2.5 personal air samples were collected from a pool of 21 elderly (65+ years of age) subjects. These subjects were residents of the 18-story retirement facility where residential monitoring was conducted. Mean daily central indoor and residential apartment concentrations were approximately 10 µg/m3. Outdoor and ambient PM2.5 concentrations averaged 22 µg/m3 with a daily range of 6.7–59.3 µg/m3. The slope of the central indoor/outdoor PM2.5 mass relationship was 0.38. The average daily ratio of PM2.5/PM10 mass co ncentrations across the measurement sites ranged from 0.73 to 0.92. Both the central indoor and mean apartment PM2.5 mass concentrations were highly correlated with the outdoor variables (r>0.94). The lack of traditionally recognized indoor sources of PM present within the facility might have accounted for the high degree of correlation observed between the variables. Results associated with the personal monitoring effort are discussed in depth in Part 2 of this article.

Comparison of PM2.5 and PM10 monitors.

An extensive PM monitoring study was conducted during the 1998 Baltimore PM Epidemiology-Exposure Study of the Elderly. One goal was to investigate the mass concentration comparability between various monitoring instrumentation located across residential indoor, residential outdoor, and ambient sites. Filter-based (24-h integrated) samplers included Federal Reference Method Monitors (PM2.5-FRMs), Personal Environmental Monitors (PEMs), Versatile Air Pollution Samplers (VAPS), and cyclone-based instruments. Tapered element oscillating microbalances (TEOMs) collected real-time data. Measurements were collected on a near-daily basis over a 28-day period during July–August, 1998. The selected monitors had individual sampling completeness percentages ranging from 64% to 100%. Quantitation limits varied from 0.2 to 5.0 µg/m3. Results from matched days indicated that mean individual PM10 and PM2.5 mass concentrations differed by less than 3 µg/m3 across the instrumentation and within each respective size fraction. PM10 and PM2.5 mass concentration regression coefficients of determination between the monitors often exceeded 0.90 with coarse (PM10–2.5) comparisons revealing coefficients typically well below 0.40. Only one of the outdoor collocated PM2.5 monitors (PEM) provided mass concentration data that were statistically different from that produced by a protoype PM2.5 FRM sampler. The PEM had a positive mass concentration bias ranging up to 18% relative to the FRM prototype.

Evaluation of the Sodium Arsenite Method for Measurement of NO2 in Ambient Air

The sodium arsenite method for measurement of nitrogen dioxide in ambient air was evaluated. The method has a constant-high collection efficiency (82%) for nitrogen dioxide, and is insensitive to normal variations in operating parameters. Nitric oxide and carbon dioxide are positive and negative interferents, respectively. The combined average effect of these interferents, over ambient levels, is a positive bias of 9.9 µg/m3. This bias, although statistically significant, is minor ( 10 % ) in relation to the ambient air standard of 100 µg NO2/m3 and does not warrant modification of the method to remove the interference.

An empirical bayes method for comparing air pollution data to air quality standards

Abstract Air pollution concentrations are often characterized by skewed distributions. For example, 24-h suspended particulate mass concentrations have historically been characterized by a 2-parameter lognormal distribution (Larsen, 1969, J. Air. Pollut. Control. Ass. 19 , 24–30). The error in making measurements of these concentrations also conforms to a distribution that is more symmetric and can usually be approximated by the normal distribution, often with the added feature that the error variation is a function of concentration level. Air pollution measurements are often used to determine status with respect to legal standards such as the National Ambient Air Quality Standards (NAAQS) (Fed. Reg. 36 , 8186, 1971; 44 , 8202, 1979). In general, these standards do not explicitly account for measurement error. Using Bayesian mathematics, a model is developed which deals with the probability of the “true” concentration exceeding the NAAQS given a measurement above the NAAQS. Several hypothetical examples are used to demonstrate that Bayesian techniques can combine both pollution history and instrument precision into a probablistic model for comparing air pollution data to NAAQS.

Lead analysis of ambient air particulates: interlaboratory evaluation of EPA lead reference method.

The evaluation of the U. S. Environmental Protection Agency’s “Reference Method for the Determination of Lead in Suspended Particulate Matter Collected from Ambient Air” among four laboratories and the quantitation by additional techniques are summarized. Coefficients of variation of 10% for the lead reference method were obtained. Optical emission spectrometry produced data that are statistically indistinguishable from data obtained from the reference atomic absorption spectrometry technique.

Effects of measurement uncertainty on air quality summary statistics

Abstract This paper examines the effects of measurement uncertainty on various summary statistics that are routinely used in air quality data analysis. Analytical approximations and computer simulation techniques are employed to illustrate and quantify how the uncertainty associated with an individual measurement results in an uncertainty for different summary statistics. Measurement uncertainty may be viewed as consisting of bias and imprecision. It is shown that even when there is no bias for individual measurements it is possible for imprecision alone to result in bias for certain commonly used summary statistics. Different types of statistics are shown to be less influenced by measurement imprecision and, consequently, a data set may be acceptable for some purpose but not for others. The desired precision of the summary statistic may be viewed as a guide in determining an acceptable level of imprecision for individual measurements.