Another possibility, Calle said, is that the study masked any protective effect on cancer by looking at all cancers combined, rather than specific cancers.
"Cancers are very different from one another, and risk factors for cancer are very different," she said. "If you looked at individual cancers you might see things that you don't see with all cancers combined."
Willett also noted that some fruits and vegetables may have an effect on some types of cancer.
"I think it is plausible that there are some components of fruits and vegetables that may modestly reduce the risk of some cancers, but lumping all fruits and vegetables together obscures the benefit," he said. "For example, we have seen evidence that a higher intake of tomato-based products may reduce the risk of prostate cancer."
In addition, the researchers found a protective association for cruciferous vegetables (such as cauliflower, cabbage, broccoli, and even mustard and collard greens, for instance) and cancer, but only in men. Whether the types of cancers occurring in men are more responsive to these types of vegetables compared to cancers in women remains to be determined.
The bottom line, Calle said, is that studying the effects of foods on disease is a very complex process. "While the data don't really indicate a reduction in risk for all cancers combined, we're not really ready to believe there's no reduction for individual cancer sites," she said. "Fruits and vegetables are healthy choices whether we can directly show this impact on all cancers combined or not."