New research out of Harvard University’s Chan School of Public Health came to the not-exactly-earth-shattering conclusion that higher intakes of sugary beverages are associated with increased risk for type 2 diabetes (T2D). Researchers also determined that even artificially sweetened beverages come with increased risk for T2D, but that replacing sugar-sweetened beverages with water, tea, or coffee results in a slight decrease in risk. However, the data upon which these findings are based have several flaws and the findings should be interpreted with caution.

The paper, “Changes in Consumption of Sugary Beverages and Artificially Sweetened Beverages and Subsequent Risk of Type 2 Diabetes: Results From Three Large Prospective U.S. Cohorts of Women and Men,” was published last month in the journal Diabetes Care. It collected information from men and women in the medical profession who had participated in large, long-term prospective cohort epidemiological surveys between 1986 and 2013 (the Nurses’ Health Study, the Nurses’ Health Study II and the Health Professionals’ Follow-up Study). Probably the single largest weakness of the information gathered is that it was based on food frequency questionnaires administered once every four years. If you have a hard time remembering what you had for lunch five days ago, imagine trying to estimate the amount of sugar-sweetened or artificially-sweetened beverages you’ve consumed over the past four years. Could you provide an estimate? Yes, probably. Would that estimate be accurate? Likely not.

Food frequency questionnaires (FFQs) are so highly flawed and the data they produce so unreliable that researchers have said they contain “fatal flaws” and have “engendered a fictional discourse” on the relationship between dietary factors and disease. They’ve gone so far as to assert that the US national dietary guidelines are based on “inadmissible” data, and that using memory-based dietary assessment methods, like FFQs, “constitutes an unscientific and significant misuse of research resources,” and that uncritical faith in the validity and value memory-based methods “has wasted significant resources and constitutes the greatest impediment to scientific progress in obesity and nutrition research.”

So, proceeding with the understanding that the data informing these findings are shaky at best, let’s proceed and see if there are useful points we can take away.

After adjusting for BMI and initial and changes in diet and lifestyle confounders, increasing total intake of sugary beverages by just half a serving per day—including those sweetened with added sugar as well as those containing 100% fruit juices—was associated with 16% increased risk for T2D during the subsequent 4 years (95% CI 1–34%). One serving was defined as 8 ounces, so an increase of just 4 ounces per day was correlated with this increased risk. Increased intake of artificially-sweetened beverages (ASBs), also by more than half a serving per day, was associated with an 18% higher risk (CI 2–36%). Finally, researchers determined that swapping out one daily serving of a sugary beverage in favor of water, coffee or tea, but not an ASB, was associated with a 2–10% lower risk. With confidence intervals that include 1 and 2 percent, we’re not exactly talking huge increases or decreases here, but we can still potentially find some takeaways.

First, these researchers should be applauded for including 100% fruit juice in their data collection. When fingers are pointed at sugar-sweetened sodas for causing all manner of health problems, fruit juice too often gets a free pass. Natural sugar is still sugar, and it can wreak havoc on blood glucose every bit as powerfully as soda, particularly when it hits the body in liquid form that doesn’t even need to be chewed or broken down. The fructose in fruit doesn’t affect blood sugar in the same way that glucose does (whether as part of a sucrose molecule or as free glucose), but owing to the way it’s metabolized in the liver, high fructose consumption over the long term may lead to weight gain, dyslipidemia and insulin resistance. People who think they’re doing themselves a favor by substituting fruit juice or a fruit smoothie for their sugar-sweetened soda may not realize it’s possible they’re consuming even more sugar than before.

Second, the researchers acknowledge that the association between ASB consumption and increased risk for T2D may be due to reverse causation—that is, people who are at higher risk for T2D or who already have it may be more inclined than the general public to consume ASBs in an effort to control their sugar intake and manage their blood glucose. As for the finding that replacing sugar-sweetened beverages (SSBs) with ASBs was not associated with reduced T2D risk, this isn’t surprising. If someone swaps out SSBs or fruit juices for ASBs but makes no other change to their diet or lifestyle, we shouldn’t expect much change in their risk for T2D. Reducing consumption of liquid sugar is never a bad thing, but in people at risk for diabetes or those who already have it, reducing total carbohydrate intake across the board may be necessary to make a dent in improving blood glucose control. For example, cutting back on juice and soda may not have much of an impact on glycemic control in people who still consume large quantities of pasta, bread, cereal, muffins, crackers, chips, and other high-starch and high-sugar foods. In a consensus report on nutrition therapy for adults with diabetes or prediabetes issued earlier this year by a panel from the American Diabetes Association, the authors noted that reducing overall carbohydrate intake has shown the most evidence for improving glycemia. (Click here to read our coverage of this report.) 

One final piece of food for thought comes from the potentially reduced risk for T2D that was associated with replacing SSBs with water, coffee, or tea. It should be noted that coffee and tea mean coffee and tea, not mildly coffee- or tea-flavored concoctions loaded with sugar and sweetened milks or creams. Many of these fancy beverages have double or triple the amount of sugar in a soda but it’s easy for patients to be unaware of their detrimental effects, partially because they’re drinks, and partially because they may be considered “discretionary” treats and aren’t factored into a day’s calorie, carb, or fat intake. An app or spreadsheet that tracks food intake might not know if one of these beverages was consumed, but the liver and pancreas sure do.