“Of the 13 nutrients that we were able to study, we found statistically reliable declines in six of the 13,” he says. Levels of other nutrients stayed roughly constant over the years.

But a big word of caution: USDA nutritionist Joanne Holden says those 1950 numbers may not be trustworthy. For one thing, measurement techniques have changed, possibly changing the results. In addition, she says, no one knows whether the vegetables measured in 1950 were an accurate sample of the American diet.