My Morning Cup of "What the Eff??" Red Meat Won't Kill You

Last week a media shit-storm erupted when a study was published from Harvard MD’s claiming that increased consumption of red meat would lead to higher mortality risk, cardiovascular injury risk, diabetes risk, the game of Risk, and any other type of risk imaginable.

If you want to read the actual study, check it out HERE.

I’m all for studies that look at specific components of diet relating to increased risks, but this study kind of jumped the shark huge in their assumption of red meat causing increased mortality, and I’d like to go through some of the big parts of why this made me facepalm for the better part of the week and weekend.

Let’s start with the obvious assumptions. saying that red meat is exclusively linked to increased mortality would mean that the researchers would have to have controlled big variables such as physical activity, total caloric intake and smoking status, right?

Nope.

They clearly labeled the percentage of participants in each quinttile of increasing red meat consumption who smoked, and showed that it increased significantly through the increasing red meat consumption, with 5.0% of the lowest quartile being active smokers and  14.5% of the highest quartile being current smokers. Physical activity went down through the quartiles of increasing consumption, going from 27.5 MET hours per week to 17.5 MET hours per week, or a reduction in activity by 36%. THAT’S A BIG DIFFERENCE!!

Caloric intake increased through the quintiles, ranging from 1202 in the lowest quintile of the nurses study to 2396 in the upper quintile of the health professionals study, or DOUBLE THE CALORIC INTAKE!  Interestingly enough, the intake of red meat went up from 0.22 servings per day in the lowest quintile to 2.36 servings per day (an increase of 10 times the amount of red meat per day), was a difference of only 2 servings a day, only resulted in an increase to the pooled hazard ratio of 1.0 up to 1.2, meaning the increase was significant, but not crazy huge, and would be more likely corellated to the relative risk of increased smoking based on the percentages of individuals who smoked.

The increased intake of one serving per day of red meat, which encompassed about 40-60% of the population at any given time, showed an increased hazard ratio of 12% for total red meat intake, 13% for unprocessed red meats (steak, pork, etc), and a 20% increase in unprocessed red meat (bacon, hot dogs, etc). The increased energy intake during this change in red meat servings intake was roughly 30% depending on which quintiles you’re looking at (a difference of only 17% comparing the highest to 4th quintile in the nurses study), and physical activity tended to decrease by 30% per increase in serving.

So a substantially increased caloric consumption and decreased activity level wasn’t the issue here, it was the increased intake of a 4 ounce serving of red meat per week that was the causative factor for people dying prematurely, is that what you’re telling me?

While I won’t argue that the numbers these researchers came up with were accurate based on the information they had, the big issue I have is with epidemiological studies looking at proving something from an inference and observational analysis instead of from a scientific trial. Sure, there’s an increased risk with increased red meat consumption, but what isn’t discussed is the other glaringly obvious features of the study participants, namely the difference in caloric intake, physical activity levels, and level of smoking. What would have been interestingly to see would be the difference in body mass intake from baseline for the five different quintiles, and see if there was an increase in mortality.

If you were looking to make a real causative analysis into whether red meat consumption caused an increase in mortality, you would have to pick groups of people who were fairly equal in body size, activity level, caloric intake, lifestyle factors, and have the only different factor be the amount of red meat they consumed, and base the results on the results of that intervention. This study looked at groups of people who were vastly different in a lot of different variables and made a conclusion about one factor that may or may not have had much of a relation to what thy were looking at.

Based on the information found in this study, saying that red meat consumption was responsible for an isolated increased risk of mortality would be like saying drunk drivers who speed and crash their cars are more likely to die if they drive a Kia versus a Subaru. THAT’S NOT THE REAL ISSUE!!!

As we increased the red meat consumption, participants also drank more, ate more processed foods, had higher BMI’s, but also had lower total cholesterol.

Wait, what?? Lower cholesterol?? Isn’t cholesterol a big risk factor to cardiovascular issues? Too bad they never talked about that in the study though, even though it was sitting right in front of their face. They could have also said eating more red meat caused a reduction in cholesterol, but that wouldn’t be nearly as convenient for a media campaign now, would it?

One of the big problems with media reporting studies is they only give you the basics, and a lot of the media reports on the information they feel is convenient or that will create a story, not necessarily what is accurate. These researchers had a few things working for them. They were MD’s from Harvard, which sounds really impressive, so they have to know what they’re talking about. A glaring headline like “Red Meat will Kill You” and all of a sudden you have the Associated Press picking up the story and broadcasting it across the world.

The data gathered on lifestyle factors were all self-reported, including how much red meat the participants ate in the previous month. I can assure you if I were to ask 99 people out of 100 how much red meat they had last month, they would all either over- or under-estimate their intake, because we don’t typically keep track of things like that. Additionally, a lot of studies have shown we underestimate our caloric intake, overestimate our physical activity, and percieve our IQ’s to be substantially higher than average, so the data being collected was somewhat biased and inaccurate.

Ever play the telephone game? Where you and a bunch of your friends sit around and one person begins by whispering in their neighbours ear something like “My favourite colour is blue,” and by the time it makes its’ way around all your friends it comes out as something like “the body is hidden behind the swingset.” A study that looks at the data from another study after the fact is kind of like playing telephone, the message is somewhat garbled.

Additionally, the relative risk of mortality was significantly higher in a lot of studies that looked at the effects of increased smoking, but that never really factored into the equations with this study, unless there was an increase in red met consumption to go along with it. Same thing with the increased risk with increased alcohol consumption, sugar consumption, decreased activity level, increased overall caloric intake, and even something like salt intake (which wasn’t a variable that was looked at), but whatever.

I like red meat. I tend to eat about 3 servings each week, 4 if we go out for dinner on the weekends. Reading a study like this is in no way going to deter me from wanting to have a nice red steak on my plate a few times a week, because I’m also going to be working out, eating good quality foods, and not smoking or drinking. That will be much more health-protective than simply skipping red meats and grabbing another pack of menthols.