You understand what she's saying. And not only is it tough to be a woman, it's tough to be a man. On prostate tests, should men screen? Operate? Oh, now don't operate. And the latest, don't even screen. What should men do?
And what about children? Treat their teeth with fluoride or not? Eat organic or not? Give them plastic bottles or not?
The bottom line is, we all care to some extent about personal health choices, and depend on the latest research to make our decisions. Doctors can help by passing on recommendations based on the research and their own filtering of the risks of one action or another. But the research can be confusing. One research study rarely drives a decision, rather, bodies of research sway medical recommendations like the one to recommend that women take hormones after menopause to preserve youth. And now, 50 years later, the recommendation that women not take hormone pills.
When hormones were first recommended for women a half a century ago, even then there were concerns about possible side effects. But doctors, women, and media surged ahead with treatment. Now, after many women have stopped hormone therapy recent studies are showing that some women benefit from hormones while for others there's limited risks. The science is slowly capable of a finer grain analysis of the issues.
The tricky part is translating the results of many research studies into public health recommendations. As this challenges doctors and those in public health, patients are also confused by what sometimes seems like an arbitrary process. The barrage of pharmaceutical ads on television is not helpful. And the barrage of "studies" reported in the press is mind-numbing. Based just on the media, it's too easy for the consumer to view each study as a separate public health recommendation, since the press presents studies not necessarily to educate but grab eyeballs and sell ads.
Take for instance the press report last week on research that people who eat fatty meals then drink coffee can raise their blood sugar -- dangerously. Science Daily published an article titled: "Got a Craving for Fast Food? Skip the Coffee, Study Suggests". The title is not very intuitive, but hundreds of news outlets explained by quoting the author of the study, who stressed the importance of the study's findings for people with diabetes and metabolic disease: "We have known for many years that people with or at risk of Type 2 diabetes should limit their caffeine intake". She continued:
"New Evidence That Drinking Coffee May Reduce the Risk of Diabetes (June 10, 2010)" (my emphasis). Aha! That story conflicts with the other two in saying that caffeine may reduce diabetes.
What Happens to Mice who Drink Coffee Instead of Water?
I'm going to call this the "The Caffeine Controversy". The latest 2011 and 2008 stories appear to agree, so lets look at the 2010 story, published in "ACS' Journal of Agricultural and Food Chemistry". I'm not an expert in caffeine physiology or diabetes, so I'm going to do some basic stuff to try to learn more about whether I should believe the study.
First, who's the publisher? Different journals have different levels of clout or respect. ACS, the American Chemical Society is a professional organization for chemists, not to be confused with the lobby group the American Chemical Council (ACC). The researchers come from Nagoya and Kinki Universities, in Japan, as well as Pokka Corporation, a drink company. 4 out of 12 researchers come fom Pokka Corporation and the coffee used in the study was "a gift". So hmmm...It's not that coffee is that expensive or that great research isn't done by corporations, but just in general, how often does corporate sponsored research show that their product is dangerous to health? But lets keep looking at this study anyway.
The actual title of the paper is "Coffee and Caffeine Ameliorate Hyperglycemia, Fatty Liver, and Inflammatory Adipocytokine Expression in Spontaneously Diabetic KK-Ay Mice", which is way more nuanced than the press title "New Evidence That Drinking Coffee May Reduce the Risk of Diabetes". The study looks at physiological markers of lab mice genetically altered to become insulin resistant. The mice were given coffee instead of water in their diet, before being tested for biochemical markers hyperglycemia and diabetes.
We could look further at the specific tests they did, their statistics, the length of the study, the amount of caffeine used, or the effects of substituting coffee for water. We could examine their specific results, for instance fasting blood sugar was statistically insignificant between both groups, so they did an insulin tolerance test that showed the desired difference. But like most people, we don't have the immediate knowledge of the idiosyncrasies of these tests, so this would take a fair amount time. So instead let's look for studies that seem from the outset to be without conflicts of interest. Not that you should ever make assumptions.
What Happens to Humans who Drink Coffee?
So far we've looked a three studies and still don't have an answer, therefore the benefits or dangers of coffee remain "a controversy". And we don't even drink coffee, so do we care? But we're really curious about this statement from the researcher in the latest 2011 study, "We have known for many years that people with or at risk of Type 2 diabetes should limit their caffeine intake."
Why are we curious? This is a significant statement. According to NIH statistics from 2011, diabetes affects 25.8 million people. This amounts to 8.3 % of the US population according to the NIH, 11% of people over 20 years old, and 26.9% of people over 65 years old. About 30% of people over 65 years old have undiagnosed diabetes.
According the National Coffee Association daily coffee drinkers make up around 50% of the US population of about 300 billion people. Obviously, there's an overlap between these two huge groups. So it would be really relevant if the advice "avoid coffee consumption" were to be added to "exercise and lose weight" to prevent diabetes? And if this is the case, why then, do at least some hospitals treating patients who are diabetic allow them to drink coffee?
We'll turn to PubMed, where lots of published science research is collected. Fortunately, other researchers have also turned to Pubmed or MEDLINE to answer this very question. Two epidemiology studies have recently found that coffee actually lowers incidence of type II diabetes. These are fairly large studies that if true would dispute the current study. Lets look briefly at them.
One group from Harvard's Departments of Nutrition and Epidemiology, Brigham and Woman's, Harvard Medical School, and Vrije Universiteit's Department of Nutrition and Health, searched MEDLINE through January 2005 and found nine cohort studies culminating 193,473 study participants. They results of all these studies show that habitual coffee consumption decreases risk of Type II diabetes (van Dam et al: "Coffee Consumption and Risk of Type 2 Diabetes A Systematic Review" JAMA. 2005;294(1):97-104. doi: 10.1001/jama.294.1.97). That's a pretty solid epidemiological finding. The group doesn't appear to have conflicts of interest.
A second group with researchers from the US, France, Australia, Netherlands and Scotland. Huxley, R. et al: "Coffee, Decaffeinated Coffee, and Tea Consumption in Relation to Incident Type 2 Diabetes Mellitus: A Systematic Review With Meta-analysis" looked at prospective studies between 1966 and 2009, 18 studies with 457,922 participants, also found an inverse relationship between diabetes and self-reported coffee, as well as tea and decaffeinated coffee drinking. This too is a solid finding. So these two studies differ from the one we're looking at, but it's fair to say that the results of epidemiological studies can differ from studies showing some metabolic influences of coffee.
Should Humans Forgo Coffee?
Of course there are more studies, in humans, in mice, epidemiological studies, and biochemical and physiological studies. For now, although it seems as though coffee may indeed alter glucose homeostasis, this may not add up to something that can be seen in epidemiological studies. It doesn't mean coffee doesn't have an effect, or isn't harmful. We could keep looking at studies if we drank coffee and wanted to make a decision about this. But circling back to the original study, we'd venture that the Ph.D student/researcher's statement, "we've known for years" that people at risk for diabetes (a third the population) shouldn't drink coffee (1/2 the population) is at best hyperbolic. More so considering that the paper's discussion section notes that one of their results may explain the "negative correlation between long- term coffee consumption and type 2 diabetes risks".
And since this was reported in hundreds of news reports, lets also look quickly at the methodology. The 2011 paper used 11 volunteers. Being that this was a controlled experiment, subjects fasted for 12 hours after going two days without coffee, exercise or alcohol. The researchers then had participants drink a "fat cocktail", which consisted of 1 gram of fat/1 kg of body weight. (I don't know what the exact fat composition of the drink was because I couldn't find the "Supplemental Table II".)
But if you were a 160 pound male (72 kg), your experimental "fat cocktail" would consist of 72 grams of fat, which amounts to 5.5 tablespoons of soybean oil (one of the ingredients used in the study); or more familiar to most people, 24 tablespoons of half & half cream; more than 3 McDonald's Double Cheeseburgers; (.pdf); or about 3 orders of large McDonald's French Fries. I don't know what you think, but this pile of food would be outside the range of and meal choice for me. The participants then waited five hours, before drinking the caffeine equivalent of 2-3 cups of coffee (5 mg/kg body weight). 1 hour later they were fed 75g of dextrose (like glucose) - about 75 grams of high-glycemic carbohydrates. By comparison, a large Coke from McDonalds has about 86g of carbohydrates and a package of sugar has about 4 grams of carbohydrates. This protocol, the fat then the sugar after a 12 hour fast, caused a physiological response in the participants. Suprise?
Crave Fast Food? Skip The Three McDonald's Cheeseburgers
"Craving Fast Food, Skip the Coffee", the title of the press report warns. But what if when "craving fast food" you just skipped the three cheeseburgers? What if you just had a small coffee in the morning, with your non-fat yogurt or your dry toast, ok maybe a pat of butter. What would that do? What if the press report for any study actually reported the real story about the research?
Or, what if the press report just included the actual title of the research? In this case the title of the research was: "An Oral Lipid Challenge and Acute Intake of Caffeinated Coffee Additively Decrease Glucose Tolerance in Healthy Men." This is a lot different than what the media reported. And while eating 6 tablespoons of soybean oil upon arising AM after a 12 hour fast might be something some people do, and indeed the results may be interesting, how does this translate to any sort of public health recommendation like the ones the authors and news is trying to make?
What if when interviewed, the lead author said, as she did in her paper 1 that while a few studies have shown glucose responses to caffeine, there is actually a "negative correlation between long- term coffee consumption and type 2 diabetes risks?" What Science Daily published instead was basically a false statement "We have known for many years that people with or at risk of Type 2 diabetes should limit their caffeine intake", which the researcher qualified (above) by advising people not to eat a lot of red meat and to drink decaf. Isn't this last advice, just common sense? But then does it follow from their study? No. It's previous research.
As a consumer of health news, it's worth reading the actual studies, or even just looking at the title, because as we showed, they often contradict the headline in the press. Secondly studies differ. Epidemiological studies where people self report, differ from other literature reviews, differ from lab mice studies, and from studies where people are attached to an IV. Different methodologies between the same type of study can yield different results.
As consumers we could try to understand all the nuance differences, but like Gail Collins, I think it's impossible. It's enough just to know that different methodologies can produce different results but that doesn't necessarily mean anything. This may sound confusing, but it isn't any more confusing then talking to multiple people about anything, from fixing the squeak in your car to whether your tie looks good.
You may rightly point out that the caffeine controversy is different than the estrogen controversy that Collins refers to. But it's really not. You have scientific research presented by the media, which is a meld of companies with vested interests, scientists, reporters biases and limitations, and doctors and clinicians. Consumers (patients) need to make sense of it all.
Tragically, people got cancer from estrogen therapy. People will die from heart attacks, obesity, diabetes. Decisions they make about coffee may influence the rate of their demise, but we don't exactly know how. The indefiniteness of research today does not help us make today's decisions.
Or does it? Is it the science or medical professions who confuse us? Doctors? Or the press and pharmaceutical companies? Or do we confuse ourselves rather than trying to understand some basic stuff about scientific publishing, press releases, the news industry, doctors, and business? In the case of our 2011 science research study on caffeine, the actual peer-reviewed published study was fairly informative about the limitations of the research. Even the title was elucidating. I haven't yet seen widespread physician's recommendations regarding the dangers of coffee. The most hyperbolic accounts in this case occurred in the press (perhaps with the help of the Ph.D. student - and where was the adviser?)
Most professions are required to take continuing education credits. If we're in charge of our health except for periodic ten minute interactions with the doctor, maybe we should be trying to understand how science, medicine and news industries work in order to take care of ourselves? I'm not talking about diagnosing ourselves. In this case, most consumers know what a somewhat healthy diet looks like, and that it doesn't involve 3 orders of large fries at McDonalds for breakfast. People know they need to exercise. The consensus of scientists and doctors is not controversial, it's simple, we need not be confused. But fruits, vegetables and exercise don't sell newspapers and pharmaceutical drugs.
----------------
1 M.-S. Beaudoin, L. E. Robinson, T. E. Graham. An Oral Lipid Challenge and Acute Intake of Caffeinated Coffee Additively Decrease Glucose Tolerance in Healthy Men. Journal of Nutrition, 2011; 141 (4): 574 DOI: 10.3945/jn.110.132761