Step I Items	Response
Confounding for BPA and obesity	•Body composition (age, ethnicity, gender, height, race);  •Weight (age, gender);  •Waist circumference (age, gender);  •Body mass index (age, ethnicity, gender, race);  •In addition, consumption of canned or packaged food and drink (“processed” food) that is also energy dense and low-nutrient (e.g., soda) is a significant confounder because food packaging is a main source of exposure to BPA.  •Co-exposures: There may be some concern for co-exposure to certain phthalates used in food packaging that have also been linked to obesity. However, phthalates are used in different types of food packaging than BPA (plastic wraps versus canned lining and polycarbonate materials). No other a priori co-exposures of particular concern are identified for general population studies. There may be some co-exposures that need to be considered in occupational studies and these should be assessed on a case by case basis if discovered.
Co-interventions	•None identified
Accuracy of the measurement of exposure to BPA (CAS# 80–05-7)	•BPA is a non-persistent compound (near 100% elimination within 24 h after oral exposure, possible longer elimination time from non-oral exposure but on order of days), so blood and urine measures only assess recent exposure. This means current exposure levels may NOT be indicative of past exposures. This is problematic for assessment of BPA as a risk factor for health outcomes that are not acute and take time to develop like obesity.  •BPA measures are variable over time in the same person (even during the same day) so methods that utilize repeated measures of exposure are preferred. Some experts on BPA exposure assessment express less concern for lack of repeated measures for NHANES data because it is a large sample survey of the general population.  •Standard analytical measures: Measurement of urine or blood by quantitative techniques such as liquid chromatography-triple quadrupole mass spectrometry (LC-MS/MS) and high-pressure liquid chromatography with tandem mass spectrometry (HPLC/MS) are preferred. Measurements made at CDC are considered high-quality.  •Measures to minimize sample contamination with BPA should be taken (e.g., glass pipettes, polypropylene plastic lab ware and sample collection materials, water blanks).  •Measures of unconjugated BPA in blood need to be very carefully considered based on extent to which investigators controlled for background exposures.  •Questionnaire or self-reported measures of BPA exposure are more problematic due to the ubiquity of exposure and lack of knowledge on all possible routes of exposure, e.g., thermal paper, certain pharmaceuticals. However, there is some support for an association between higher urine/blood levels of BPA and higher reported use of BPA-containing food packaging (e.g., canned food consumption) or handling of BPA-containing thermal paper (cashiers) so questionnaire data that assess these types of exposure sources may have some utility in assessing longer-term time trends in exposure.
Accuracy of the measurement of outcome of obesity	•Body Composition: Dual-energy X-Ray absorptiometry, triceps skinfold thickness, subscapular skinfold thickness, suprailiac skinfold thickness  •Measured waist circumference  •Body mass index  •Measured weight *Obesity typically develops relatively slowly over time so preferred follow-up times after start of exposure would be on the order of several months to years.