March 23, 2009
Increased Mortality With Higher Red Meat Consumption
Red meat and processed meat probably cut life expectancy.
Individuals who eat more red meat and processed meat appear to have a modestly increased risk of death from all causes and also from cancer or heart disease over a 10-year period, according to a report in the March 23 issue of Archives of Internal Medicine, one of the JAMA/Archives journals. In contrast, a higher intake of white meat appeared to be associated with a slightly decreased risk for overall death and cancer death.
"Meat intake varies substantially around the world, but the impact of consuming higher levels of meat in relation to chronic disease mortality [death] is ambiguous," the authors write as background information in the article.
Rashmi Sinha, Ph.D., and colleagues at the National Cancer Institute, Rockville, Md., assessed the association between meat intake and risk of death among more than 500,000 individuals who were part of the National Institutes of Health-AARP Diet and Health Study. Participants, who were between 50 and 71 years old when the study began in 1995, provided demographic information and completed a food frequency questionnaire to estimate their intake of white, red and processed meats. They were then followed for 10 years through Social Security Administration Death Master File and National Death Index databases.
During the follow-up period, 47,976 men and 23,276 women died. The one-fifth of men and women who ate the most red meat (a median or midpoint of 62.5 grams per 1,000 calories per day) had a higher risk for overall death, death from heart disease and death from cancer than the one-fifth of men and women who ate the least red meat (a median of 9.8 grams per 1,000 calories per day), as did the one-fifth of men and women who ate the most vs. the least amount of processed meat (a median of 22.6 grams vs. 1.6 grams per 1,000 calories per day).
When comparing the one-fifth of participants who ate the most white meat to the one-fifth who ate the least white meat, those with high white meat intake had a slightly lower risk for total death, death from cancer and death from causes other than heart disease or cancer.
"For overall mortality, 11 percent of deaths in men and 16 percent of deaths in women could be prevented if people decreased their red meat consumption to the level of intake in the first quintile [one-fifth]. The impact on cardiovascular disease mortality was an 11 percent decrease in men and a 21 percent decrease in women if the red meat consumption was decreased to the amount consumed by individuals in the first quintile," the authors write. "For women eating processed meat at the first quintile level, the decrease in cardiovascular disease mortality was approximately 20 percent."
Processed meat appears to have a more potent effect as compared to red meat. So hot dogs and salami pose substantial risks. Why is that?
White meat does not appear to cut life expectancy.
Among women, those who ate the most red meat were 36 percent more likely to die for any reason, 20 percent more likely to die of cancer and 50 percent more likely to die of heart disease. Men who ate the most meat were 31 percent more likely to die for any reason, 22 percent more likely to die of cancer and 27 percent more likely to die of heart disease.
In contrast, those who consumed the most white meat were about 8 percent less likely to die during the study period than those who ate the least, the researchers found. Poultry contains more unsaturated fat, which improves cholesterol levels, and fish contains omega-3 fatty acids, which are believed to help reduce the risk of heart disease.
I find the white meat result a little surprising. I wonder what the effect is of eating the dark meat in turkey. I also wonder whether consumption of more omega 3 fatty acids will reduce the harm from eating lots of red meat.
Here's the full research report.
Correlation or association?
Perhaps unhealthy people eat more read meat because they are unhealthy.
The ones who ate the most meat are the ones who ate the most carbohydrates. The increased mortality came from inflammation caused by excess insulin. Like most nutrition studies by the NIH, they meant to prove that meat was bad, but they confirmed what numerous studies have already proved-that carbohydrates are the killer.
I organically manage my cattle on grass and flax seed to enhance the Omega-3 content of the milk from my dairy and in the meat we produce. Animals that eat a natural diet rather than packed with grain in a CAFO are healthier to eat, so if you want to be critical of beef, criticize CAFO beef which has essentially no Omega-3 content. A grass fed animal, fed grain in a CAFO for 30 days consumes virtually all of their body's Omega-3 before slaughter. Not all red meat is the same. Eat grass fed animals only. It goes for poultry too.
Self-report of food eaten. Worthless.
This is another in a long string of sensationalized data mining 'research' papers. Aside from the problem with self reports (see succinct comment by ywegwe5), there are sample problems (AARP members not representative of population, only about 15% return on the questionnaire, etc.), and most importantly, the groups are not comparable. The highest quintile red meat group had a higher BMI, were less educated, smoked much more, had much less exercise, ate more total calories (including much less fruit, much more fat and saturated fat, and much less fiber), and took fewer vitamins and supplements than the lowest quintile group. They were simply much less healthy.
The authors tried to statistically manipulate the data to 'remove' the differences in the groups by building a model that 'controlled' for the differences, but the model still found that people who ate red meat were much more likely to die from injury and sudden death (p=.01). The relative risk was a 26% increase, greater than the increased risk of cancer death (24%) and comparable to the cardiovascular increase (27%). Unless someone can come up with a plausible explanation why eating red meat will increase the probability of having a car wreck or falling under a bus, this pretty much points to a sample difference not related to meat consumption, and invalidates the conclusions.
BTW, buried in the results is this gem: According to their model, if you eat a lot of red meat you SHOULD smoke--it reduces the risk of death from all causes!
From Table 1 of the paper, it is shown that the group (Q5) with the highest red meat intake consumed 119 g/kcal of all meat combined. The data of total amount of all meat for the group with the highest white meat intake is not shown but my estimation based on the reported data for this group is 69 g/kcal.
So, it seems that people who mostly eat white meat consumed about 2 fold less total meat than people who eat red meat.
People with highest intake of white meat have lower risk of death than those with lowest intake, as reported. But those with low intake of white meat actually consume more red meat and total meat in general (table 1).
Bottom line, the data overall shows a link between total amount of meat and mortality. The color of meat is irrelevant. I have made this observation before on the original papers by Willett linking red meat with colon cancer. see my book chapter in Cancer Epigenetics: http://www.amazon.com/Cancer-Epigenetics-Trygve-Tollefsbol/dp/1420045792/ref=sr_1_1?ie=UTF8&s=books&qid=1226425804&sr=1-1
I wish that the authors could make a follow up revision and change the conclusion "Red and processed meat intakes were associated with modest increases in total mortality, cancer mortality, and cardiovascular disease mortality." to "Higher meat intake were associated with modest increases in total mortality, cancer mortality, and cardiovascular disease mortality."
These "studies" are so tiring. I don't eat processed meat (hotdogs, etc) at all, but eat some hambuger. the study's conclusions with respect to percentages are worthless.
Has/will/can a privacy-scrubbed version of the dataset be made public? Notwithstanding the fact that this is self-reported diet data and does not have a good way to control for income which is important, there is still a lot of data in these half-million surveys that have been matched with mortality data.
There may be some component of public funding, but even if not, just in the interests of science and to verify the original results, it would be good for other people to be able to look for other correlations between diet and health. It may also be possible to match surveys with addresses that could add median-income or home value by zip code to the data, which would go far towards controlling for the fact that people with more financial resources are generally healthier.
It seems to me that a lot of valuable information may be going to waste if it is only used for this one study that may leave something to be desired in how it was designed.