The link between diabetes and cancer is a fast-developing area of research. A new study claims that people with diabetes may be at a greater risk of developing head and neck cancers.
In a study of over 89,000 people in Taiwan, researchers found that head and neck cancer (HNC) types were 1.47 times more common in newly diagnosed patients with diabetes than in a control group. The study published in JAMA Otolaryngology Head and Neck Surgery, noted that all the people who developed cancer and “controls” were well-matched for certain health risks: obesity and heart disease, and other factors, like sex and age.1
Oral cancer, oropharyngeal cancer, and nasopharyngeal carcinoma all had higher incidence rates in patients diagnosed with diabetes than the control group. Lead author of the study, Yung-Song Lin, MD, of the Taipei Medical University in Taiwan, said they were surprised that there was not only a prevalence of oral and oropharyngeal cancers but of nasopharyngeal carcinoma (NPC), as well.
"There must be a different underlying mechanism [that] needs to be explored as the cause of increasing HNC in patients with diabetes," Dr. Lin said. Dr. Lin said that the connection between diabetes and cancer can be complex, though.
"Both diabetes and cancer are multifactorial diseases," Dr. Lin said, "Several potential pathophysiological pathways can contribute to their interdependence."
While Dr. Lin's research suggests a connection between diabetes and HNC, others in diabetes care and research feel that the study's conclusions are misleading.
"It takes away the emphasis on far more important things," said Ralph Audehm, MD, a general practitioner and clinical director based in Victoria, Australia, and an advisory panel member for Diabetes Australia, a nationally syndicated nonprofit organization.
"Type 2 diabetes patients do have higher rates of cancer," Dr. Audehm said, but the study's conclusions "make (HNC) seem far more important," than the other medical problems diabetes patients face.
Dr. Audehm also questions whether the data left out other information, like patients' quality of health coverage and additional physiological factors.
Another study published in Cancer Epidemiology, Biomarkers & Prevention found a weak association, at 1.09, between diabetes and HNC after analyzing data pooled from 6,448 cases and 13,747 controls.2
Stephen M. Schwartz, PhD, of the Fred Hutchinson Cancer Research Center in Seattle, Washington, took part in this study and commented on the recent research done in Taiwan, pointing out important differences between the two studies.
"Unlike our studies, this Taiwan study includes nasopharyngeal carcinoma [NPC] with its definition of head and neck cancers," Dr. Schwartz said.
NPC is more common in Taiwan than in the United States, Dr. Schwartz said, and while both studies used administrative records from health insurance claims, the Taiwanese study differed in that it left out smoking as a factor.
Taiwanese HNC cases also can be linked to the chewing of betel quid, an addictive nut chewed commonly in Asia. The differing mix of exposures to the data populations could contribute to the differences between the two studies, Dr. Schwartz said.
In the US study, Dr. Schwartz found a 1.59 association between people with diabetes who did not smoke, and 0.96 among diabetics that used tobacco—essentially no discernible association. Dr. Schwartz said these findings present an interesting debate in exploring the connection between HNC patients who have diabetes and smoke.
"Many people believe that it is among non-smokers (or non-users of tobacco and alcohol) that one should expect to find associations with something like diabetes," Dr. Schwartz said. "But others would say the opposite: that since tobacco and/or alcohol figure prominently in the causation of so many (HNC) cases, factors like diabetes most likely are interacting with them."
The consensus is changing that the causes of HNC are fully understood, Dr. Schwartz said. "This view is changing for HNC, because some HNC (especially those of the oropharynx) seem to be due primarily to human papillomavirus infection (HPV), which demonstrates to us that even when tobacco and alcohol use are dominant causes, there is room for other factors to be acting, particularly in subsets of the disease," Dr. Schwartz said.
"The strengths of the study include the size of the database cohort, and the fact that the database includes the majority of the population in Taiwan," noted Tamara L. Wexler, MD, PhD. However, "limitations include those common to any database-based retrospective cohort study—for example, a lack of contact with subjects and reliance on data as entered. Particularly relevant to a study looking at HNC is the inability to directly assess tobacco history, as well as other exposures (alcohol, betel, EBV or HPV) noted by the authors. As stated, association does not denote causality, and it may be that other health behaviors influenced the results," she added.
"In addition, there is no indication of the time course (prior to diagnosis) or subsequent control of diabetes in the diabetic population. Being able to determine whether the risk of HNC was related to the degree of hyperglycemic exposure would strengthen the author's association," said Dr. Wexler, an endocrinologist and internist, and a Clinical Associate at Massachusetts General Hospital in Boston.
"In terms of the relevance to other populations, HNC is also associated with environmental factors, some of which may be particular to Taiwan. While this article suggests the importance of assessing patients with type 2 diabetes for head and neck cancers (i.e. on physical examination), further data is needed before drawing additional conclusions," she noted.
The study design makes drawing any conclusions difficult. Instead, it should prompt further evaluation, such as with a prospective study incorporating known environmental contributors to HNC, as well as patient interviews.
The new study featured in JAMA Otolaryngology Head and Neck Surgery was supported through a grant from The Taipei Medical University and Chi Mei Medical Center Research Fund.