In a large population screened for cervical cancer with state-of-the art techniques, overweight and obese women had a lower rate of precancer but an elevated risk of invasive cervical cancer. The findings are likely due to the failure to detect precancer in women with a high body mass index (BMI), the researchers said.
Obese women had the lowest 5-year risk of precancer at 0.51% (95% CI 0.48-0.54) versus 0.73% (95% CI 0.70-0.76) for their normal or underweight counterparts (P trend <0.001). Overweight women also had a lower rate: 0.61% (95% CI 0.58-0.64). But obese women had the highest 5-year cancer risk at 0.083% (95% CI 0.072-0.096) compared with 0.056% (95% CI 0.048-0.066) in normal or underweight women (P trend <0.001).
In the study, published online in the Journal of Clinical Oncology, the investigators suggested that detecting precancerous lesions in these weight categories may be challenging with routine screening, but better techniques and equipment to ensure adequate cervical biopsies and visualization could potentially reduce the incidence of cervical cancer in heavy women.
Led by epidemiologist Megan A. Clarke, PhD, MHS, of the National Cancer Institute in Bethesda, MD, the retrospective cohort study looked at 944,227 women ages 30 to 64 who underwent cytology and human papillomavirus (HPV) DNA testing at Kaiser Permanente Northern California during 2003-2015.
Using values recorded in the preceding 5 years, the researchers categorized BMI as normal/underweight (<25), overweight (25 to <30), or obese (>30). The patients were followed for a mean of 4.4 years, and most were ages 30 to 49; close to 40% were white, although heavy women were more likely to be black or Hispanic. About 6.3% were HPV-positive.
Overall, approximately 20% of cervical cancers were attributable to overweight or obesity.
A total of 4,489 women (0.48%) developed precursor cervical intraepithelial neoplasia 3 or adenocarcinoma in situ, and of these, 1,998 were classified as normal/underweight (44.4%); 1,339 as overweight (29.8%); and 1,152 as obese (25.6%). A total of 490 women (0.05%) were diagnosed with cancer, and of these, 149 were classified as normal/underweight (30.4%); 154 as overweight (31.4%); and 187 as obese (38.2%).
Patients thus appeared to have a lower risk of having clinical precursors of cervical cancer but a higher risk of invasive cancer with increasing BMI. The results were consistent in subgroups defined by age (30-49 versus 50-64), HPV status (positive versus negative), and histologic subtype (glandular versus squamous).
“Any risk factor associated with increased cervical cancer risk would also be expected to increase precancer risk, and likewise, a decreased precancer risk should lead to a corresponding decreased risk of cancer, Clarke and associates wrote. “Thus, the apparent paradoxical association of excess BMI with decreased risk of precancer, but increased risk of cancer is likely related to underdiagnosis of cervical precancer in overweight and obese patients who were screened for cervical cancer.”
Asked for his perspective, Don S. Dizon, MD, of Brown University and Lifespan Cancer Institute in Providence, RI, who was part of the study, said he agreed with the authors that the results are likely due to underdetection rather than a direct causal effect of obesity. He also agreed that screening heavy women is challenging from a practical standpoint: “Can the patient get on the table and into the stirrups? Can you find an appropriate speculum with adequate tensile stretch that will allow for direct visualization of the cervix?”
The pressure effect of obesity may lead to sampling errors, he explained: “Some Pap tests may not be done because direct visualization is not possible.”
Dizon also pointed out that the database study did not indicate what type of practitioner did the screening – i.e., a busy family doctor with 15 minutes per patient who does maybe one a week, or a gynecologist who does five or six a day? – and that variable can lead to sampling errors.
Clarke and colleagues noted that obesity-related screening or imaging problems have also emerged in prostate cancer, resulting in underdiagnosis, and have also been noted in radiology reports.
As for cause and effect, previous studies on the association of obesity and increased cervical cancer incidence and mortality have not been consistent, and most studies have lacked screening information or have not analyzed precancer endpoints separately, the researchers noted. And they stressed that while HPV vaccination is promising, screening will remain a preventive mainstay.
MedPage Today has reported on previous findings that obesity is a barrier to screening for cervical and breast cancers.
Clarke et al said that in light of the worldwide obesity epidemic, the findings highlight the need for greater awareness and the development of specific clinical recommendations for screening overweight and obese women.
“Whether existing equipment (e.g., speculum size) and procedures are adequate for visualization and sampling of the cervix in overweight and obese women requires further evaluation,” they wrote.
Dizon said that better tools will always be helpful, “but the important issue is that we need to provide good primary care for all Americans regardless of body size. And if you needed another reason why healthy eating habits, lifestyle, and activity should be stressed more within the U.S. general population, this may actually be it: If you have a higher BMI, we may not be able to give the proper surveillance to potentially catch cancer before it becomes invasive.”
The researchers noted that the study was limited by an inability to relate specific false-negative co-test and/or colposcopy results to cancer diagnosis, since the exact timing of cancer development versus detection was not always known. In addition, BMI data were missing for about 13% of the cohort. Furthermore, since women tend to gain weight over time, using 5-year median BMI at enrollment may have led to misclassification, thereby potentially underestimating a true association between obesity and cancer. Finally, there was no information about patient comorbidities.
The study was supported by the Intramural Research Program of the National Cancer Institute.
Clarke reported having no conflicts of interest. Two co-authors reported financial relationships with Teva, Bayer HealthCare, Roche, and Becton Dickinson.
Robert Jasmer, MD Associate Clinical Professor of Medicine, University of California, San Francisco and Dorothy Caputo, MA, BSN, RN, Nurse Planner