Low-carbohydrate diets have increased in popularity, and there is good evidence to show that they result in short-term weight loss. But there is less data available with respect to their long-term health effects.

A recent meta-analysis aimed to analyze the correlation between carbohydrate intake and mortality, or risk of death. You’ve probably heard about it, and you’ve probably heard that, based on this study, low-carb diets decrease life span. But what does the study actually show?

The authors analyzed a study of 15,428 adults aged 45 to 64 years, from four different U.S. communities. The subjects completed a dietary questionnaire between the years of 1987 and 1989 and were then followed for a mean of 25 years. The authors also used a meta-analysis to compare this study to seven other similar studies.

The authors found a U-shaped curve such that mortality rates were higher below 40 percent of calories from carbohydrate and above 70 percent from carbohydrate. They found the lowest risk of mortality to lie at 50 to 55 percent of calories from carbohydrate.

The authors went deeper and also analyzed mortality risk for different types of low-carbohydrate diets. They found that, for those who ate most of their calories from animal sources, mortality rate was, in fact, higher. Those on low-carbohydrate diets who ate most of their calories from vegetable sources, however, actually had a lower mortality rate. Thus, it seems that carbohydrate percentage isn’t linked with mortality at all. The real factor is the type of foods consumed.

Remember: those who ate low-carbohydrate diets high in animal proteins and fats did have an increased risk of mortality. But the authors found that these subjects were actually more likely to be overweight or obese, exercise less, smoke cigarettes, and have diabetes. Any of these alone could be the cause of their increased risk of mortality, let alone the whole group of risk factors. This is known as the “healthy user bias” and is a common problem in research. If people are more likely to eat meat, they are more likely to have unhealthy lifestyles, because of the long-term vilification of meat. It is thus impossible to separate out the consequence of meat versus that of their unhealthy lifestyle.

So this article can’t really say anything about the mortality risk of a low-carbohydrate diet that includes animal protein in the context of a healthy lifestyle.

Further, the quality of a low-carb diet is important, and the study was not able to analyze the quality of the animal fats and proteins consumed. On a low-carb diet, vegetables and fiber are important and healthful, and they were not necessarily consumed by the subjects in the study. Thus, a low-carbohydrate diet has not been shown to increase risk of death. An appropriate low-carbohydrate diet will include quality fats and protein (from both plants and animals), minimal processed foods, ample vegetables, nuts, seeds, and fiber, and a complete nutrient intake.