A large study published Monday that looked at the dietary patterns of more than 100,000 Americans discovered an unexpected link between high consumption of citrus — specifically whole grapefruit and orange juice — and risk of melanoma, the most dangerous form of skin cancer.

Researchers found that 1,840 of the study participants developed melanoma and that those who had a serving of citrus fruit or juice 1.6 times daily had a 36 percent higher risk of the cancer as compared with those who consume it less than two times a week. A serving was defined as half a grapefruit, one orange or a 6 ounce glass of juice.

Writing in the Journal of Clinical Oncology, the authors theorize that the link may be due to high levels of something calledfurocoumarinsfound in citrus fruit. These substances are produced by plants as a defense mechanism and arephotoactive, meaning that their toxicity is enhanced by ultraviolet radiation. They have been known to cause skin to be more sensitive to sunlight.

The findings were independent of age, and otherfactors includingcigarette smoking, alcohol intake and other lifestyle factors linked to cancer.

ShaoweiWu, the lead study author and a postdoctoral research fellow at the Warren Alpert Medical School at Brown University, said that in a statement that the findings suggest that "those who consume a lot of grapefruit and/or orange juice should be particularly careful to avoid prolonged sun exposure."

The American Society of Clinical Oncology called the findings "intriguing" but said that it's far too soon for any changes to dietary recommendations about grapefruit and oranges.

In a commentary accompanying the study, Marianne Berwick from the department of internal medicine at the University of New Mexico pointed out a number of weaknesses in the study.

For one, she pointed out that the study population_ which was comprised of about 64,000 nurses and 42,000 men in the health profession — is not representative of the general population. Because of their background they might have been able identify cancerous lesions that aresmaller resultingin a higher number of diagnoses that in a different population. Berwick said that as health professionals they "might be expected to pay more attention to unusual lesions."

There was also what she called a "major inconsistency" in the findings on risk and the form of the citrus the participants in the study consumed. "[R]iskfrom grapefruit was from the wholefruit butnot the juice and risk from orange was from the juice but not the whole fruit," Berwick wrote.

Marion Nestle, a professor of nutrition, food studies, and public health at New York University, expressed skepticism about the study's findings in an interview. "The number of cases is small and the citrus risks are not cleanly dose-related," she said. "I'd worry much more about alcohol and cigarettes."

The citrus study follows closely on the heels of another melanoma study — this one published in the June issue of JAMA Internal Medicine linking Viagra use with melanoma. The study found that men who used the erection-enhancing drug were at 84 percent higher risk of developing melanoma. David Penwick, executive editor of Harvard Men's Health Watch, cautioned that the finds showed the study found an association — not causation. "1) This study does not show that Viagra causes skin cancer. Instead, it shows that in a large group of men, those who said they used Viagra ended up being diagnosed more often with melanoma than those who didn't use this drug. The study shows a connection, not a cause. 2). Even if Viagra does promote melanoma, the absolute increase is small," he wrote.

Nestle quipped, "If you take Viagra, better not take it with citrus or alcohol, and most definitely do not have a cigarette afterwards."