Researchers identify surprising link between poor diet and lower cancer risk

By Naveed Saleh, MD, MS
Published March 31, 2021

Key Takeaways

The risk of ovarian cancer is lower in women who consume diets with a higher glycemic load (GL) and glycemic index (GI), according to the surprising results of a prospective study published in the Journal of Nutrition.

“Ovarian cancer is the fifth most common cause of cancer death among women in the US, yet few modifiable risk factors have been established,” the authors wrote. “Diets high in GI and GL have been linked to several cancers, but epidemiologic studies of ovarian cancer have yielded inconsistent results.”

In this high-powered study, researchers mined data from the Prostate, Lung, Colorectal, and Ovarian cancer screening cohort, with GI and GL calculated using validated food-frequency questionnaires (FFQs) in women aged 60 to 74 years without a history of cancer and with both ovaries present. The researchers analyzed data separately for participants who completed dietary questionnaires at baseline or later during the study.

In participants who completed dietary questionnaires at baseline, the researchers identified 181 cases of ovarian cancer among 24,633 women, with a median follow-up duration of 12.1 years. In 42,410 women who completed dietary questionnaires later during the study, 211 cases of ovarian cancer were identified, with a median follow-up duration of 8.9 years. 

After taking covariates into consideration, including age at dietary questionnaire completion, year of randomization, year of questionnaire, study center, and oral contraceptive use, the team found the risk of ovarian cancer risk dropped by 43% (HR: 0.57; 95% CI: 0.37, 0.88) in the highest GL quartile vs the lowest GI quartile among participants who completed dietary questionnaires at baseline. In those who completed dietary questionnaires later during the study, the risk dropped by 38% in the highest GI quartile vs the lowest GI quartile (HR: 0.62; 95% CI: 0.42, 1.00). 

The finding that glycemic measures and risk of ovarian cancer are inversely related ran contrary to the researchers’ initial hypothesis. Of note, the authors observed no differences in GI between those participants with a history of diabetes who completed the dietary questionnaires at baseline vs those without a history of the disease. However, GI was slightly higher in women with diabetes who completed the dietary questionnaires later in the study. Nevertheless, compensating for diabetes status did not result in changes in observed relationships.

The authors had some insight into the surprising results of the study. 

“Given the current evidence and complex relations between various components of diet, exercise, and body composition, GI and GL could be considered more as components of a healthy lifestyle rather than as isolated modifiable risk factors. In this study, however, it is possible that assessment of diet did not reflect [their] usual diet during the critical window of exposure.”

The authors noted that there are few observational studies relating to GI or GL and ovarian cancer risk, with most such research performed outside of the United States. These international studies don’t necessarily focus on Western diets, thus making their findings less generalizable. 

Nonetheless, several retrospective case-control studies demonstrated an increased risk of ovarian cancer related to higher GL vs this risk in healthy female controls. Furthermore, a prospective Canadian study demonstrated increased risk with higher dietary GL. In the United States, however, results from the American Association of Retired Persons (AARP) Diet and Health Study suggested an inverse association between GL and ovarian cancer risk.

Looking forward, the authors noted that research needs to be done to understand whether the observed associations are true or a fluke. They also recommended that further studies focus on the mechanisms of these associations. 

These future studies should span more heterogeneous diet intake among participants as well, and they could include more cases to explore relationships in samples.

Share with emailShare to FacebookShare to LinkedInShare to Twitter