Who's Got the Bacon?
by Larry Creswell, M.D.
Last month a group of investigators from the Harvard School of Public Health, headed by An Pan, PhD, published a report in the Archives of Internal Medicine, entitled “Red Meat Consumption and Mortality”.
The report received a great deal of media attention. But how much of what was reported in the news should concern you?
The report focused on a study of 37,698 men from the long-standing Health Professionals Follow-up Study (HPFS) and 83,644 women from the long-standing Nurses’ Health Study (NHS) who were tracked for a period of 22 to 28 years. The subjects answered questionnaires every two to four years that characterized their typical food consumption.
The authors made three primary observations:
- Observation #1: The amount of red meat consumption (with unprocessed meat and even more so with processed meat) had an influence on mortality rate due to cardiovascular disease (CVD) and to cancer. An increase of just one serving of red meat per day was associated with a 13% to 20% increased risk of mortality. And, moreover, there was no “safe” amount of red meat consumption.
- Observation #2: By statistical means, the investigators estimated that substitution of one serving per day of alternative foods (fish, poultry, nuts, etc.) for one serving of red meat was associated with a reduction of 7% to 19% in the risk of mortality depending upon which foods were substituted.
- Observation #3: The authors estimated that 9.3% of deaths in men and 7.6% of deaths in women could be prevented if all individuals consumed fewer than a half serving of red meat per day.
As mentioned above, the report received a great deal of attention in the media. A couple examples of sensational headlines included:
All red meat is risky, a study finds
- Los Angeles Times
Scientists Warn ‘Red Meat Can be Lethal’
A week’s worth of print and television reports made me recall the children’s story about Chicken Little, Henny Penny, and the sky falling.
All of the findings may well be true, but...
- Pecking order - In scientific circles there is a pecking order to studies. In the case of this red meat report, we’re talking about an observational study. In this type of study, investigators gather a bunch of people and make some observations over time. This is at the low end of the pecking order because a variety of things can go wrong with data collection and interpretation. For reference, at the top of the pecking order is the randomized, controlled trial (RCT), where an experiment is conducted. As a pertinent example, you might gather a group of subjects, feed them a well-prescribed diet, control for all other important variables, and measure the mortality rate over time. Observational studies often generate hypotheses that are later tested in RCT’s. That’s the natural order of scientific study.
The current study took advantage of the NHS and HPFS which were established in 1976 and 1986, respectively. These long-standing studies enrolled middle-aged health professionals who were thought to be motivated and reliable. Every two years, questionnaires on a variety of topics are sent to the subjects and various investigators cull through the questionnaires to see what can be learned.
The NHS and HPFS are two of the largest and most important cohorts of Americans ever studied, so it’s important not to overlook the good that can come from observational studies. In the NHS alone, we’ve learned a great deal about topics such as the relationship between obesity and breast cancer and between oral contraceptives and coronary heart disease. But as an example of where things can go awry, data from the NHS among other observational studies famously suggested that hormone replacement therapy (HRT) in women reduced the risk of heart disease. But we now know from more recent RCT’s that, in fact, HRT actually increases the risk of heart disease substantially.
There are just a multitude of methodologic problems with this type of study. In this red meat study, you might imagine how difficult it would be to describe, by questionnaire, your typical meat consumption over the previous two-year period. But more importantly, there are always variables that the investigators can’t control for, and it’s these unmeasured variables that sometimes skew the results. As just one quirky example from the current study, it turns out that the individuals who ate the most red meat actually had the lowest cholesterol levels. That’s obviously counter-intuitive, since we often hypothesize that red meat consumption leads to higher cholesterol levels which, in turn, leads to heart disease. As one critic put it recently, you didn’t see the media (or investigators) report that, “Red Meat Reduces Cholesterol Levels!”
- Magnitude of risk - When the authors report a 13% to 20% increased risk of death attributable to just a one portion per day increased consumption of red meat, it sounds bad. But even if the difference is true, is the difference meaningful?
During the 22-year period in the HPFS, in a total of 758,524 person-years, the overall death rate was 1.18% and the mortality rates due to CVD or cancer were only 0.36% and 0.41%, respectively. In absolute terms, only about 4 per 1,000 in that period died of CVD and approximately another 4 per 1,000 died from cancer. The absolute risk of dying is obviously very small. And even if that risk were increased by 13% to 20% because of eating an additional serving per day of red meat, we’re only talking about approximately 4.5 per 1,000 individuals dying from CVD or about 4.5 per 1,000 individuals dying from cancer.
The same is true in the NHS, where for a 28-year period and 2,199,892 person-years, the overall death rate was 0.68% and the mortality rates due to CVD and cancer were only 0.15% and 0.29%, respectively. In absolute terms, less than 2 per 1,000 in that period died of CVD and approximately another 3 per 1,000 died from cancer. And again, even if that risk were increased by 13% to 20% because of eating an additional serving per day of red meat, we’re still talking about approximately 2.4 per 1,000 individuals dying from CVD and about 3.6 per 1,000 individuals dying from cancer.
Getting back to observation #3 above, recall that the authors estimated that about 10% of deaths could be eliminated if all individuals consumed less than a half serving of red meat per day. So doing the math for the men in the HPFS, the overall mortality rate would decrease from 1.18% to 1.06% and for the women in the NHS, the overall mortality rate would decrease from 0.68% to 0.62%.
I often link this notion of meaningful differences to the lottery -- which has also been in the news this past week. If you purchase one MegaMillions Lottery ticket, your chances of getting all six numbers correct is about 1 in 176 million. If you purchase two tickets, your chances double! But is that meaningful? After all, 1 in 176 million is close to zero. But so is 1 in 88 million.
- Happy medium - I’ve found that in medicine -- and in biological systems in general -- there is often a happy medium. It’s true in this study, too. The investigators divided the subjects into five groups, or quintiles, according to how much red meat they consumed. The first quintile ate the least and the fifth quintile ate the most. As it turns out, if we look only at the raw, unadjusted data, the mortality rate was lowest not in the first quintile, but in the third -- for both men and women. So, for those among us who champion the consumption of no red meat, beware the possibility that less is not always more.
So where does that leave us?
I wouldn’t rush to overhaul your red meat consumption based on this study, particularly if you’re counting on living longer as a result. I don’t doubt that scientists will continue to work on this issue and that we’ll learn more about the relationship between red meat (and other dietary items) and longevity. I don’t usually offer nutritional advice except for my patients, but I’ve taken a kind eye in recent years to Gordo’s suggestion of a real food diet. Simply eat real food (lean proteins, fruits and vegetables) and avoid processed foods, pasta and refined sugar late in the day. Nobody is going to argue for more bacon or hot dogs, but I’ll bet that a moderate intake of unprocessed red meat is probably okay.
Larry Creswell, M.D., is a cardiac surgeon and Associate Professor of Surgery at the University of Mississippi Medical Center in Jackson, Mississippi. In addition to his regular column on Endurance Corner, he maintains The Athlete's Heart blog to offer information about athletes and heart disease in an informal way and to encourage exchange and discussion that will help athletes build a heart-healthier lifestyle. You can contact him at firstname.lastname@example.org.