In 2012, the Harvard School of Public Health published a study linking red meat consumption to increased risk of total, cardiovascular, and cancer mortality. The research joined a slew of other studies that point to associations between red meat and health problems.

Dr. Mathieu Lalonde, an organic chemist at the Harvard Department of Chemistry and Chemical Biology, is questioning these findings. At the “Eat Red Meat, Save the Planet” event, hosted by the Harvard Food Law Society, Dr. Lalonde criticized the research methods and conclusions of the study, led by Dr. Frank Hu. According to Lalonde, there currently is no research that can adequately link red meat consumption and increased risks for mortality.

The Harvard study, conducted by doctors from the Harvard Medical School and the Harvard School of Public Health as well as researchers from other hospitals and institutes, observed 37,698 men and 83,644 women from the Health Professionals Follow-Up Study and the Nurses’ Health Study. Dietary data was collected from participants every four years from 1984 to 2006 using a food frequency questionnaire (FFQ) and biennial follow-up questionnaires collecting lifestyle and medical information such as weight, exercise, medications, smoking, and family history of illness. The FFQ asked participants to estimate their usual intake of different types of food for the previous year. In determining unprocessed red meat consumption, the study included beef, pork, and lamb as a main dish or part of a sandwich or mixed dish, as well as hamburger. Processed red meat included bacon, hot dogs, sausage, salami, and bologna, as well as other processed red meats.

Dr. Lalonde was skeptical about these methods. He pointed out how difficult it would be to recall your food intake for the past year—most of us can’t remember what we ate for dinner three nights ago. Thousands of participants had to be excluded for reporting physiologically impossible calorie intake levels, which leads one to wonder how accurately participants were really able to gage their consumption.

He also questioned the idea that hamburger was included under the unprocessed meat category. Most hamburgers offered at restaurants or fast food chains could probably be described as processed meat.

In addition, people with medical conditions were found to over-report their meat consumption. Participants with higher meat consumption were more likely to have higher BMI, exercise less, smoke, and drink alcohol. The study claimed to control for total energy intake, whole grain, fruit, and vegetable consumption as well as other potentially confounding variables such as age, BMI, race, smoking, alcohol intake, exercise, medication, and family history of illness. But Dr. Lalonde was not convinced.

Diet and mortality are both enormous, multivariate categories. Finding a link between consumption of a certain food and resulting health outcomes may be too large of a question.

The major issue, according to Lalonde, is that there are simply too many confounding factors to be able to ascertain a relationship. Red meat consumption is likely to be part of a broader dietary and lifestyle pattern. Diet and mortality are both enormous, multivariate categories. Finding a link between consumption of a certain food and resulting health outcomes may be too large of a question, he suggested.

The Harvard study concluded that red meat intake has an adverse effect on mortality risk. The researchers suggested that this could be due to saturated fat and cholesterol levels in red meat. They also implicated the sodium and nitrites in processed meats and potential carcinogens and heme iron found in red meat.

Lalonde argues that, as of yet, there have been no random controlled trials or animal studies to support that red meat increases risk for mortality. As he analyzed other studies that claimed red meat has negative health consequences, he expressed frustration with the fact that the causal statements promulgated by such research has gained public momentum.

The question of red meat and mortality deserves further research. But if there’s one thing we’ve learned, it’s that such research will always be subject to the difficulties posed by confounding variables. While it may be easier to determine the nutritional constituents of red meat, it’s harder to make connections between these properties and eventual health outcomes.

While it is valuable to recognize links between diet and health, it’s also important to consider them in their broader contexts and to take into account the many factors involved.