Limits...
Exposure Classification and Temporal Variability in Urinary Bisphenol A Concentrations among Couples in Utah--The HOPE Study.

Cox KJ, Porucznik CA, Anderson DJ, Brozek EM, Szczotka KM, Bailey NM, Wilkins DG, Stanford JB - Environ. Health Perspect. (2015)

Bottom Line: For the medium tertile, specificity reached 0.87 with 10 repeat samples, but even with 11 samples, sensitivity and PPV did not exceed 0.36.Five repeat samples, among males, yielded sensitivity and PPV values ≥ 0.75 for the high and low tertiles, but, similar to females, classification for the medium tertile was less accurate.Cox KJ, Porucznik CA, Anderson DJ, Brozek EM, Szczotka KM, Bailey NM, Wilkins DG, Stanford JB. 2016.

View Article: PubMed Central - PubMed

Affiliation: Office of Cooperative Reproductive Health, Division of Public Health, University of Utah, Salt Lake City, Utah, USA.

ABSTRACT

Background: Bisphenol A (BPA) is an endocrine disruptor and potential reproductive toxicant, but results of epidemiologic studies have been mixed and have been criticized for inadequate exposure assessment that often relies on a single measurement.

Objective: Our goal was to describe the distribution of BPA concentrations in serial urinary specimens, assess temporal variability, and provide estimates of exposure classification when randomly selected samples are used to predict average exposure.

Methods: We collected and analyzed 2,614 urine specimens from 83 Utah couples beginning in 2012. Female participants collected daily first-morning urine specimens during one to two menstrual cycles and male partners collected specimens during the woman's fertile window for each cycle. We measured urinary BPA concentrations and calculated geometric means (GM) for each cycle, characterized the distribution of observed values and temporal variability using intraclass correlation coefficients, and performed surrogate category analyses to determine how well repeat samples could classify exposure.

Results: The GM urine BPA concentration was 2.78 ng/mL among males and 2.44 ng/mL among females. BPA had a high degree of variability among both males (ICC = 0.18; 95% CI: 0.11, 0.26) and females (ICC = 0.11; 95% CI: 0.08, 0.16). Based on our more stringent surrogate category analysis, to reach proportions ≥ 0.80 for sensitivity, specificity, and positive predictive value (PPV) among females, 6 and 10 repeat samples for the high and low tertiles, respectively, were required. For the medium tertile, specificity reached 0.87 with 10 repeat samples, but even with 11 samples, sensitivity and PPV did not exceed 0.36. Five repeat samples, among males, yielded sensitivity and PPV values ≥ 0.75 for the high and low tertiles, but, similar to females, classification for the medium tertile was less accurate.

Conclusion: Repeated urinary specimens are required to characterize typical BPA exposure.

Citation: Cox KJ, Porucznik CA, Anderson DJ, Brozek EM, Szczotka KM, Bailey NM, Wilkins DG, Stanford JB. 2016. Exposure classification and temporal variability in urinary bisphenol A concentrations among couples in Utah-the HOPE study. Environ Health Perspect 124:498-506; http://dx.doi.org/10.1289/ehp.1509752.

No MeSH data available.


Related in: MedlinePlus

Surrogate category analysis among females, analysis 2, for PPV, sensitivity, and specificity. The displayed values are the minimum (bottom whisker), 25th percentile (bottom of box), median (line in box), mean (diamond), 75th percentile (top of box), and maximum (top whisker). The horizontal line at 0.8 is a reference line. The x-axis represents the number of repeated samples. Tertiles were defined as follows: low (< 1.76 ng/mL), medium (≥ 1.76 ng/mL to < 3.33 ng/mL), high (≥ 3.33 ng/mL). For corresponding numeric data, see Supplemental Material, Table S2.
© Copyright Policy - public-domain
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4829981&req=5

f2: Surrogate category analysis among females, analysis 2, for PPV, sensitivity, and specificity. The displayed values are the minimum (bottom whisker), 25th percentile (bottom of box), median (line in box), mean (diamond), 75th percentile (top of box), and maximum (top whisker). The horizontal line at 0.8 is a reference line. The x-axis represents the number of repeated samples. Tertiles were defined as follows: low (< 1.76 ng/mL), medium (≥ 1.76 ng/mL to < 3.33 ng/mL), high (≥ 3.33 ng/mL). For corresponding numeric data, see Supplemental Material, Table S2.

Mentions: Female, analysis 2. When the repeat samples used in the surrogate tertile were not used in the calculation of the “true” GM, the sensitivity, specificity, and PPV dropped compared with those of analysis 1 (Figure 2; for corresponding numeric data, see Supplemental Material, Table S2). When only 1 sample was used, the sensitivity in all three exposure tertiles had medians < 0.60 (low, 0.57; medium, 0.36; high, 0.55). Six repeat samples were sufficient to exceed a median proportion of 0.80 for both sensitivity and PPV in the high exposure tertile. Specificity remained much the same as in analysis 1, and 3 samples correctly classified both the low and high tertiles with a median proportion > 0.80. Even with 11 repeat samples, median sensitivity and PPV in the medium tertile was ≤ 0.36. This is likely attributable to participants’ having samples with a mixture of both high and low BPA concentrations, which makes categorization less consistent for the medium tertile, although specificity is 0.88 (95% CI: 0.88, 0.94) with 11 samples.


Exposure Classification and Temporal Variability in Urinary Bisphenol A Concentrations among Couples in Utah--The HOPE Study.

Cox KJ, Porucznik CA, Anderson DJ, Brozek EM, Szczotka KM, Bailey NM, Wilkins DG, Stanford JB - Environ. Health Perspect. (2015)

Surrogate category analysis among females, analysis 2, for PPV, sensitivity, and specificity. The displayed values are the minimum (bottom whisker), 25th percentile (bottom of box), median (line in box), mean (diamond), 75th percentile (top of box), and maximum (top whisker). The horizontal line at 0.8 is a reference line. The x-axis represents the number of repeated samples. Tertiles were defined as follows: low (< 1.76 ng/mL), medium (≥ 1.76 ng/mL to < 3.33 ng/mL), high (≥ 3.33 ng/mL). For corresponding numeric data, see Supplemental Material, Table S2.
© Copyright Policy - public-domain
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4829981&req=5

f2: Surrogate category analysis among females, analysis 2, for PPV, sensitivity, and specificity. The displayed values are the minimum (bottom whisker), 25th percentile (bottom of box), median (line in box), mean (diamond), 75th percentile (top of box), and maximum (top whisker). The horizontal line at 0.8 is a reference line. The x-axis represents the number of repeated samples. Tertiles were defined as follows: low (< 1.76 ng/mL), medium (≥ 1.76 ng/mL to < 3.33 ng/mL), high (≥ 3.33 ng/mL). For corresponding numeric data, see Supplemental Material, Table S2.
Mentions: Female, analysis 2. When the repeat samples used in the surrogate tertile were not used in the calculation of the “true” GM, the sensitivity, specificity, and PPV dropped compared with those of analysis 1 (Figure 2; for corresponding numeric data, see Supplemental Material, Table S2). When only 1 sample was used, the sensitivity in all three exposure tertiles had medians < 0.60 (low, 0.57; medium, 0.36; high, 0.55). Six repeat samples were sufficient to exceed a median proportion of 0.80 for both sensitivity and PPV in the high exposure tertile. Specificity remained much the same as in analysis 1, and 3 samples correctly classified both the low and high tertiles with a median proportion > 0.80. Even with 11 repeat samples, median sensitivity and PPV in the medium tertile was ≤ 0.36. This is likely attributable to participants’ having samples with a mixture of both high and low BPA concentrations, which makes categorization less consistent for the medium tertile, although specificity is 0.88 (95% CI: 0.88, 0.94) with 11 samples.

Bottom Line: For the medium tertile, specificity reached 0.87 with 10 repeat samples, but even with 11 samples, sensitivity and PPV did not exceed 0.36.Five repeat samples, among males, yielded sensitivity and PPV values ≥ 0.75 for the high and low tertiles, but, similar to females, classification for the medium tertile was less accurate.Cox KJ, Porucznik CA, Anderson DJ, Brozek EM, Szczotka KM, Bailey NM, Wilkins DG, Stanford JB. 2016.

View Article: PubMed Central - PubMed

Affiliation: Office of Cooperative Reproductive Health, Division of Public Health, University of Utah, Salt Lake City, Utah, USA.

ABSTRACT

Background: Bisphenol A (BPA) is an endocrine disruptor and potential reproductive toxicant, but results of epidemiologic studies have been mixed and have been criticized for inadequate exposure assessment that often relies on a single measurement.

Objective: Our goal was to describe the distribution of BPA concentrations in serial urinary specimens, assess temporal variability, and provide estimates of exposure classification when randomly selected samples are used to predict average exposure.

Methods: We collected and analyzed 2,614 urine specimens from 83 Utah couples beginning in 2012. Female participants collected daily first-morning urine specimens during one to two menstrual cycles and male partners collected specimens during the woman's fertile window for each cycle. We measured urinary BPA concentrations and calculated geometric means (GM) for each cycle, characterized the distribution of observed values and temporal variability using intraclass correlation coefficients, and performed surrogate category analyses to determine how well repeat samples could classify exposure.

Results: The GM urine BPA concentration was 2.78 ng/mL among males and 2.44 ng/mL among females. BPA had a high degree of variability among both males (ICC = 0.18; 95% CI: 0.11, 0.26) and females (ICC = 0.11; 95% CI: 0.08, 0.16). Based on our more stringent surrogate category analysis, to reach proportions ≥ 0.80 for sensitivity, specificity, and positive predictive value (PPV) among females, 6 and 10 repeat samples for the high and low tertiles, respectively, were required. For the medium tertile, specificity reached 0.87 with 10 repeat samples, but even with 11 samples, sensitivity and PPV did not exceed 0.36. Five repeat samples, among males, yielded sensitivity and PPV values ≥ 0.75 for the high and low tertiles, but, similar to females, classification for the medium tertile was less accurate.

Conclusion: Repeated urinary specimens are required to characterize typical BPA exposure.

Citation: Cox KJ, Porucznik CA, Anderson DJ, Brozek EM, Szczotka KM, Bailey NM, Wilkins DG, Stanford JB. 2016. Exposure classification and temporal variability in urinary bisphenol A concentrations among couples in Utah-the HOPE study. Environ Health Perspect 124:498-506; http://dx.doi.org/10.1289/ehp.1509752.

No MeSH data available.


Related in: MedlinePlus