Junkfood Science: The fabled links between weight and prostate cancer

December 27, 2006

The fabled links between weight and prostate cancer

The media is getting a jump start on the traditional New Year’s kick-off to the dieting season. News reports are telling men that losing weight can lower their risk for prostate cancer. “Here's another reason for men to avoid packing on extra pounds over the holidays,” according to the Associated Press.

“This study is the first to link weight loss to a reduced risk of this disease...and adds to increasing evidence of the importance of maintaining a healthy weight throughout adult life,” said the American Cancer Society.

The study behind these news stories was by researchers with the American Cancer Society and Duke University School of Medicine who examined data on 5,232 men diagnosed with prostate cancer between 1992 and June 20, 2003. They were looking for associations between prostate cancer and weight status.

Oh, you heard it was a study of nearly 70,000 men? That’s the Trojan Number — named after the mythical ploy of the Greeks who infiltrated the city of Troy inside a giant wooden horse. As Dr. John Brignell, author of Sorry, Wrong Number! explains, it’s one of the strategies used by researchers to inflate the size of a study so it sounds more impressive when the actual number of people with the condition being studied is significantly less. In this case, only 7% of the men actually had prostate cancer and the remaining 64,759 men were “mere bystanders.” When you find a Trojan Number, cautions Dr. Brignell, “these studies are often part of a large data dredge, in which many combinations of condition and potential cause are covered, so that the inevitable coincidental excesses can be identified and claimed as significant.”

In this case, the data was culled from the Cancer Prevention Study II Nutrition Cohort, a subgroup of the Cancer Prevention Study II. This controversial database on 1.2 million Americans was created by American Cancer Society volunteers who recruited friends, neighbors and family members to complete lifestyle questionnaires in 1982. Another questionnaire asking them their height, weight, diet, and lifestyle factors was mailed to more than 84,000 men and 97,000 women living in states with cancer registries in 1992, when they were enrolled in this subgroup.

The actual findings of this study were decidedly different from what we’re hearing:

The researchers found no association between the men’s body mass index (BMI), or the men’s change in weight, and their overall risk for prostate cancer.

In fact, looking at the data, the obese men had about 10% lower overall incidence of prostate cancers and the risk steadily dropped as BMIs increased.

These held true even after the researchers tried in their statistical analyses to control for a number of differences between the "obese" men and “normal” weight men that could point to other possible factors:

·Black men were two times more likely to be obese than slim. (African Americans have a two-fold higher mortality rate from prostate cancers than white men, according to the U.S. Preventive Services Task Force.)

·Former smokers were 24% more likely to be obese than slim.

·The obese men in this cohort exercised 4 times less than the thinner men.

Most prostate cancers do not kill men diagnosed with them — while American men have a 15% lifetime risk of being diagnosed with prostate cancer, they have only a 3% lifetime risk of dying from the disease and 90% of deaths are in men over 65 years of age, according to the USPSTF. Most of the men in this study, about 90%, had PSA screening tests which can detect cancers four or more years before they would be picked up symptomatically. According to the USPSTF’s review of the medical evidence, “screening may detect cancers that appear clinically significant based on size and tumor grade, but which would not have progressed to clinical symptoms during the patient’s lifetime.”

Closely conforming to these facts, only about 5% of the prostate cancers in this study were metastatic, fatal cancers; 80% were nonmetastatic, low-grade. The remaining 662 cases were nonmetastatic high grade cancers, which they called “aggressive.”

If being fat played a role in prostate cancers, as is being suggested, risks should increase as body weights increase. They don’t. Even risks for metastatic and the nonmetastatic high grade cancers did not increase with increasing body weights. In fact, all of the statistically-derived relative risk numbers for associations between prostate cancers and body weight were so nominal they have no practical relevance. [See Have your steak and enjoy it too! for an explanation of tenable relative risks.]

With that caveat in mind, even the trivial changes in relative risks (remember these numbers are not a man’s actual risk, it is the percentage change of their actual 3% lifetime risk) reported in this study show that being fat is irrelevant and losing weight is unwarranted.

The news focused on the higher risk among “obese” men as compared to “normal” weight men for “aggressive” forms of prostate cancers, so let’s look at those numbers. There was no correlation with body weights. The 17% increased relative risk for “obese” men was actually lower than less heavy men in the “overweight” category with BMIs 27.5 to 30. And the slightly “overweight” men with BMIs of 25 to 27.5 had a 13% lower risk than the “normal” weight men!

And losing weight was even less related to lower risks for prostate cancers. The men losing 11 or more pounds had a 10% lower risk — and losing more (> 21 pounds) didn’t reduce their risks any further. That was the same reduced risk seen in men gaining more than 21 pounds! There were no differences in risks among the men who’d gained 6 to 20 pounds, were weight stable, or who’d lost 6-10 pounds.

And while the news accurately reported that the men who lost 6 to 20 pounds was associated with a 17% lower risk for the most deadly, metastatic forms of cancer — what wasn’t reported is that the men who gained 10 pounds had a 20% lower risk! And the men were gained 20 pounds still had a lower risk than the men who had "watched their weight" and not gained an ounce.

But this study’s findings are not surprising in light of the body of evidence, especially the evidence on real people. Any lingering concerns that being fat could increase risks for prostate cancer will be allayed by simply looking at the latest cancer statistics from the U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Cancer Institute, and the Surveillance, Epidemiology, and End Results Program. While obesity rates are apparently rising, prostate cancer deaths have dramatically dropped over the past ten years by one-third.

So, this study actually provided no evidence for any of the things being attributed to it: It didn’t show that being fat raised a man’s risk for prostate cancer. It didn’t show that losing weight lowered a man’s risk for prostate cancer. It didn’t show that maintaining a “healthy” weight imparted any benefits. You may be surprised by how radically different the actual findings in this study are from the spin and speculations coming from the American Cancer Society. But those familiar with this special interest group and its history of going to any length to hype obesity fears won’t be. Its corporate sponsors include the largest pharmaceutical companies with drugs for weight loss and treating “obesity-related” conditions. And Weight Watchers is the founding sponsor of the American Cancer Society’s annual Great American Weigh-In held in cities across the nation, “whose primary goal is," according to WW, "to make people aware of the link between being overweight and the risk for many diseases, especially several forms of cancer.”

© Sandy Szwarc 2006

Bookmark and Share