Note to our followers: Our nearly 13-year run of daily publication of new content on came to a close at the end of 2018. Publisher Gary Schwitzer and other contributors may post new articles periodically. But all of the 6,000+ articles we have published contain lessons to help you improve your critical thinking about health care interventions. And those will be still be alive on the site for a couple of years.
Read Original Release

More subterfuge on subgroup analysis … this time from a major cancer organization

Among nonsmokers who had diabetes, those who took the diabetes drug metformin had a decrease in lung cancer risk, according to a study in Cancer Prevention Research, a journal of the American Association for Cancer Research, by Lori Sakoda, PhD, MPH, research scientist at the Kaiser Permanente Division of Research in Oakland, California.

Some laboratory studies and a number of observational studies suggest that metformin may prevent cancer, but the data from human studies, however, are conflicting, explained Sakoda. The researchers conducted this study to further clarify the association between metformin use and lung cancer risk.

Sakoda and colleagues conducted a retrospective cohort study of 47,351 diabetic patients (54 percent men), 40 years or older, who completed a health-related survey between 1994 and 1996. Information on their diabetes medications was collected from electronic pharmacy records. About 46 percent of them were “ever-users” of metformin, defined as those who filled two or more prescriptions within a six-month period.

During 15 years of follow-up, 747 patients were diagnosed with lung cancer. Of them, 80 were nonsmokers, and 203 were current smokers.

Metformin use was not associated with lower lung cancer risk overall; however, the risk was 43 percent lower among diabetic patients who had never smoked, and the risk appeared to decrease with longer use. Nonsmokers who used metformin for five years or longer had a 52 percent reduction in lung cancer risk, but this finding was not statistically significant.

Metformin use for five or more years was associated with a 31 percent decrease in the risk for adenocarcinoma, the most common type of lung cancer diagnosed in nonsmokers, and an 82 percent increase in the risk for small-cell carcinoma, a type of lung cancer often diagnosed in smokers, but neither of these findings were statistically significant.

In an interview, Sakoda said, “Metformin use was not associated with lung cancer risk when we looked at all patients with diabetes. However, our results suggest that risk might differ by smoking history, with metformin decreasing risk among nonsmokers and increasing risk among current smokers. Our results suggesting that the risk associated with metformin might differ by smoking history were unexpected. Additional large, well-conducted studies are needed to clarify whether metformin may be used to prevent lung or other cancers, particularly in specific subpopulations, such as nonsmokers.”


This study was funded by the National Institutes of Health. Sakoda declares no conflicts of interest. Assiamira Ferrara, Charles Quesenberry Jr., and Laurel Habel, coauthors on this study, have received research funding from Takeda to Kaiser Foundation Research Institute for a study of pioglitazone and cancer and from Sanofi through a subcontract from University of North Carolina to Kaiser Foundation Research Institute for a study of insulin glargine and cancer. Habel has received additional research funding from Genentech to Kaiser Foundation Research Institute for a study of HER2-positive breast cancer, including risk of cardiotoxicity following trastuzumab.

Follow us: Cancer Research Catalyst; Twitter @AACR; and Facebook

For AACR information, visit Fast Facts

About the American Association for Cancer Research

Founded in 1907, the American Association for Cancer Research (AACR) is the world’s oldest and largest professional organization dedicated to advancing cancer research and its mission to prevent and cure cancer. AACR membership includes more than 33,000 laboratory, translational, and clinical researchers; population scientists; other health care professionals; and cancer advocates residing in 101 countries. The AACR marshals the full spectrum of expertise of the cancer community to accelerate progress in the prevention, biology, diagnosis, and treatment of cancer by annually convening more than 20 conferences and educational workshops, the largest of which is the AACR Annual Meeting with over 18,000 attendees. In addition, the AACR publishes eight peer-reviewed scientific journals and a magazine for cancer survivors, patients, and their caregivers. The AACR funds meritorious research directly as well as in cooperation with numerous cancer organizations. As the Scientific Partner of Stand Up To Cancer, the AACR provides expert peer review, grants administration, and scientific oversight of team science and individual grants in cancer research that have the potential for near-term patient benefit. The AACR actively communicates with legislators and policymakers about the value of cancer research and related biomedical science in saving lives from cancer. For more information about the AACR, visit


3 Star

Metformin may lower lung cancer risk in diabetic nonsmokers

Our Review Summary

iStock_000033186898_SmallThis release reports on an NIH-funded study to determine if the use of the drug metformin by diabetic patients may affect their risk of developing lung cancer. The overall answer is that it does not, but the release takes five paragraphs to say that. Instead, it focuses on a very small subset of patients who had never smoked and whose lung cancer risk appeared to be lower than never-smokers who didn’t take metformin. By trumpeting findings from that subgroup — and burying the main results of the study, which were negative — this release misrepresents the research findings and misleads its readers.

We’d suggest that the findings of this study probably did not warrant the spotlight provided by a news release. But if a news release is to be issued, it should lead with the results for the primary outcome and give second billing to provocative but preliminary findings from subgroups. While not as egregious as another example of subgroup “spin” that we recently reported on, this news release will certainly give readers the impression that there’s something special about metformin when it comes to preventing cancer in people with diabetes, when that hasn’t been established with any certainty.


Why This Matters

Diabetes is a chronic disease requiring long-term use of drugs such as metformin. Some research suggests that metformin may decrease the risk of lung cancer — making the results of this study an important addition to the scientific literature. Unfortunately, the release focuses almost entirely on the results seen in a small subgroup of patients. As such, it doesn’t accurately reflect the totality of the research in question. Subgroup findings can be included, but should never supersede the main outcome.


Does the news release adequately discuss the costs of the intervention?

Not Satisfactory

There is no discussion of the cost of metformin, which is available as a generic and is comparatively cheap. Nevertheless, since diabetes often requires a lifetime of drug use, costs of metformin versus other drugs would be useful to readers, especially in cases such as this when the drug under study is suspected of providing additional benefits or bringing negative side effects.

Does the news release adequately quantify the benefits of the treatment/test/product/procedure?

Not Satisfactory

We appreciate that the release attempts to quantify the reduction in risk associated with metformin use. However, there’s a problem with the release’s description of these results. The research paper points to a statistically significant reduction in lung cancer risk only among diabetic patients who have never smoked. But the release generalizes this finding to all “nonsmokers.” The release doesn’t delineate between people who have never smoked and those who did smoke but quit. The benefits appear to apply only to the group of never smokers. This is an important distinction.

Does the news release adequately explain/quantify the harms of the intervention?

Not Satisfactory

There is no mention in the release about potential harms that may arise with the administration of metformin. The drug is known to produce harmful side effects in more than 10 percent of people taking it, largely affecting the stomach or bowel such as vomiting, diarrhea, and loss of appetite.

Does the news release seem to grasp the quality of the evidence?

Not Satisfactory

Our concerns about the treatment of evidence start with the headline: “Metformin may lower lung cancer risk in diabetic nonsmokers.” Since this is an observational study that cannot prove cause and effect — only associations — we think the use of an active verb here (“may lower risk”) is inappropriate. The release should have said that metformin is “linked with” or  “associated with” lower risk, but not that it “lowered” risk. We offer a primer on this issue for health writers here.

The release does provide considerable data derived from the actual research paper, compared to many such releases, but then points to the lack of statistical significance for that data.  Releases are intended to simplify and clarify research so that general readers can understand the importance of the findings.  In this case, the release centers on a minor finding that applies to only a small subset of individuals in the study while mentioning the larger overall findings in passing. (The finding of “decreased risk” is based on just 80 people in the study who developed lung cancer despite never smoking.) That overall finding — that metformin generally has no effect on lung cancer risk — hardly qualifies as news. The decision to do a news release on this study at all is questionable.

Does the news release commit disease-mongering?


The release does not commit disease-mongering.

Does the news release identify funding sources & disclose conflicts of interest?


The release clearly identifies the funding source for the research and also adequately discloses conflicts among the authors of the paper.

Does the news release compare the new approach with existing alternatives?

Not Satisfactory

The release investigates the possibility that metformin may decrease the risk of lung cancer in people who never smoked. This would have been an opportune place to add that not smoking — or quitting smoking, if you are a smoker — is probably the most important thing one can do to reduce the risk of lung cancer.

Does the news release establish the availability of the treatment/test/product/procedure?


The story does not directly address the availability of metformin. However, it’s apparent that the drug is available, since thousands of people in the study have been taking it for years.

Does the news release establish the true novelty of the approach?


The release does reference earlier research that offered conflicting findings and supports the need for this larger study.

Does the news release include unjustifiable, sensational language, including in the quotes of researchers?


The release is cautious in its choice and use of language in explaining the current research.  It is troubling, however, that a small aspect of the research’s findings is the main emphasis of the release, while the research paper itself keeps that aspect in proper perspective.

Total Score: 5 of 10 Satisfactory


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.