Canned Soup Consumption and Urinary Bisphenol A: A Randomized Crossover Trial FREE

To the Editor: Human exposure to bisphenol A (BPA) is widespread.¹ In adults, urinary BPA concentrations are positively associated with cardiovascular disease and diabetes.² Exposure occurs primarily through the diet.³ Bisphenol A has been quantified in many canned goods,^4- 5 where it is present as a by-product of interior epoxy coatings used to prevent corrosion. We hypothesized that canned soup consumption would increase urinary BPA concentrations relative to fresh soup consumption.

In 2010, we recruited Harvard School of Public Health (HSPH) student and staff volunteers (aged >18 years) via informational fliers and e-mail. The study was approved by the HSPH institutional review board. Written informed consent was obtained.

We used a randomized, single-blinded, 2 × 2 crossover design. For the first 5-day period (October 25-29, 2010), one group consumed a 12-ounce serving of fresh soup (prepared without canned ingredients) daily between 12:15 and 2 PM; the other group consumed a 12-ounce serving of canned soup (from 18.5-ounce Progresso brand) per the same schedule. After a 2-day washout, treatment assignments were reversed (November 1-5, 2010). We used 5 different varieties of vegetarian soup per treatment and repeated them in the same order the following week. Participants were not restricted in their consumption of other foods.

Donation of spot urine samples took place between 3 and 6 PM on the fourth and fifth days of each phase. Urine was collected in polyethylene containers and stored in polypropylene cryogenic vials. When urine was donated both days, combined urine samples were created to minimize within-person variation. The specific gravity (SG) of the urine was measured using a handheld refractometer; total (free plus conjugated species) urinary BPA concentration was measured at the Centers for Disease Control and Prevention using online solid-phase extraction coupled to isotope dilution high-performance liquid chromatography–tandem mass spectrometry.⁶

Urinary BPA concentrations were adjusted for dilution by multiplying values (in μg/L) by [(1.014 − 1)/(SG − 1)]. In the absence of a period effect, we used a paired t test to examine mean absolute change in SG-adjusted untransformed urinary BPA concentrations (BPA_canned − BPA_fresh) because differences were normally distributed, and we calculated confidence intervals for the mean using SAS version 9.2. Statistical tests were 2-sided and performed at the .05 level of significance.

Of 84 volunteers, 75 (89%) completed the study. Median age was 27 years and 51 (68%) were female; median treatment adherence was 100% (Table). Bisphenol A was detected in 77% (n = 58) of samples after fresh soup consumption and 100% (n = 75) of samples after canned soup consumption. The SG-adjusted geometric mean concentration of BPA was 1.1 μg/L (95% CI, 0.9-1.4 μg/L) after fresh soup consumption (unadjusted: 0.9 μg/L; 95% CI, 0.7-1.2 μg/L) and 20.8 μg/L (95% CI, 17.9-24.1 μg/L) after canned soup consumption (unadjusted: 17.5 μg/L; 95% CI, 14.1-21.8 μg/L). Stratification by treatment sequence revealed similar values (Figure). Following canned soup consumption, SG-adjusted urinary BPA concentrations were, on average, 22.5 μg/L higher (95% CI, 19.6-25.5 μg/L) than those measured after a week of fresh soup consumption (P < .001), representing a 1221% increase.

Consumption of 1 serving of canned soup daily over 5 days was associated with a more than 1000% increase in urinary BPA. Generalizability is limited due to selection of participants from 1 school and testing of a single soup brand; however, generalizability to canned goods with similar BPA content is expected. The increase in urinary BPA concentrations following canned soup consumption is likely a transient peak of yet uncertain duration. The effect of such intermittent elevations in urinary BPA concentration is unknown. The absolute urinary BPA concentrations observed following canned soup consumption are among the most extreme reported in a nonoccupational setting. For comparison, the 95th percentile unadjusted urinary BPA in the 2007-2008 National Health and Examination Survey was 13.0 μg/L (95% CI, 10.0-15.4 μg/L).¹ The observed increase in urinary BPA concentrations following canned soup consumption, even if not sustained, may be important, especially in light of available or proposed alternatives to epoxy resins linings for most canned goods.