Even the methodology for acquisition of the self-reported data used in this study has changed over the years. > In NHANES I (1971–1975) and NHANES II (1976–198

Even the methodology for acquisition of the self-reported data used in this study has changed over the years. > In NHANES I (1971–1975) and NHANES II (1976–1980), in-person interviews were used to obtain self-reported dietary information via a 24-h dietary recall questionnaire that assessed food and beverage intake for weekdays only. In NHANES III (1988–1994), dietary information was obtained through a self-reported 24-h dietary recall using a computer-assisted, automated, interactive method for any day of the week. In NHANES 1999–2002, a multiple-pass computer-assisted dietary interview format was used to collect detailed self-reported information about all foods and beverages that were consumed the day prior to the in-person interview (weekday or weekend). In NHANES 2003–2008, 24-h self-reported dietary recalls were performed twice (3–10 days apart) using an automated multiple pass method. I'm highly suspect of this one. This seems like an interesting avenue of exploration, but this data seems too weak to make broad claims from.