It’s common knowledge that protein is the most satiating macronutrient, and there is bountiful research to support this conclusion. Yet, most all of this research has one major caveat: the content of two macronutrients is varied while the third is held constant. After all, you can’t change the protein content of a meal without changing its fat or carbohydrate content, at least not if you want calories to remain equal.

The problem with this approach is that is prevents definitive conclusions about what caused what. Was it the increase in protein that reduced appetite, or was it the simultaneous reduction in carbohydrates or fats? We can of course make sound conclusions based on the literature as a whole, which is how we arrived at our current belief that protein is the most satiating macronutrient. But the keystone study that shows without a doubt that protein is king when you want to rule over appetite has only recently been published.

I’m talking about the work of Dougkas and Östman from Lund University, Sweden. These researchers used an elaborate study design to compare the satiating effects of all three macronutrients simultaneously in a group of 36 young, healthy men and women.

The participants came to the research lab on seven occasions after an overnight fast, with at least one week separating the visits. Upon arrival, they completed a 24-hour food, physical activity, sleep, and mood questionnaire to assess any potential factors that could influence food intake. If anything atypical was observed, the study visit was rescheduled for the following week. Additionally, female participants completed a menstrual cycle questionnaire and report the first date of their last menstrual period at each study visit. All this matters because is greatly helps reduce the likelihood that some life event would confound the results.

After the questionnaires, the participants had some baseline measurements done (appetite, mood, and well-being questionnaires, as well as blood sampling) and were then provided with 1 of 7 liquid test meals served chilled in opaque cups with a lid and straw to minimize visual comparisons. Measurements were repeated over the subsequent 3.5 hours. Energy intake was assessed by an unrestricted pasta meal provided at the 3.5-hour mark.

The liquid test meals

All seven of the liquid test meals were as identical as possible despite having different macronutrient compositions. They provided 500 kcal for men and 400 kcal for women and had the same volume and therefore energy density. Moreover, the flavor, texture, and sensorial properties of the preloads were matched as closely as possible and tested in a pilot study using a panel of 20 individuals.

It is worth mentioning now that the study participant ratings of the test beverages were not significantly different. This is combination with the blinding strongly reduces the likelihood that something other than the macronutrient composition influenced any outcomes.

In order to have sufficient statistical power to detect differences between the individual macronutrients, the ranges used were quite large (figure 1). The first manipulated factor was protein, which was set at 9% (low protein; LP) or 40% (high protein; HP) of calories. The second factor was the ratio of carbohydrate to fat, which set at 0.44 (low carb; LC) or 3.56 (high carb; HC). Additionally, a center-point beverage with moderate protein (25%; MP) and a 2.0 ratio of carbohydrate to fat (MC) was consumed three times to establish experimental error and thus make possible an estimation of reproducibility.

Effects of drinking

The first set of analyses that the researchers performed compared each beverage to one another. Overall, the beverages performed similar to one another without significant differences when it came to appetite scoring. For instance, the center-piece beverage and the LP/HC beverage did not score significantly different from any other beverage in any category. Also, none of the beverages differed with regard to hunger and desire to eat. However, the HP/HC beverage scored with significantly lower total appetite and prospective food intake and significantly greater fullness than the LP/LC beverage.

Despite the above differences between some of the beverages, average energy intake during the subsequent pasta lunch meal was not significantly different between conditions. However, this finding should be taken with a grain of salt, as a significant sex effect was observed whereby men ate more than women. Accordingly, conducting the main analysis with all data could have led to notable variation that masks any real effects.

For the second analysis, the researchers then aggregated the data from all the beverages so as to provide macronutrient comparisons independent of one another. Upon doing so, it was shown that increasing protein is associated with reduced appetite, hunger, desire to eat, and prospective food intake. Increasing the carbohydrate to fat ratio was also associated with reduced appetite and desire to eat, albeit to a lesser extent than protein, as well as increased fullness (which protein didn’t effect).

This same double analysis was also applied to glycemic and hormonal responses. Blood glucose was significantly greater in the LP/HC than in any other condition, whereas it was significantly lower in the HP/LC than any other condition. Insulin was significantly lowest in the LP/LC condition, followed by the HP/LC condition, and similar among the HC conditions. GLP-1 was significantly greatest in the HP/LC condition and similar among all others. Ghrelin was unaffected by the type of beverage.

Looking to the macronutrients themselves, the results are somewhat to be expected. Reducing carbohydrates and increasing protein were both associated with lower blood glucose. Although reducing carbohydrates was also associated with reduced insulin, increasing protein was associated with increased insulin. However, for both variables the effect from manipulating carbohydrates was stronger than that from manipulating protein. Increasing protein and reducing carbohydrates was associated with increased GLP-1, with protein having a stronger effect. Finally, increased protein tended to reduce ghrelin.

Protein is the satiety king

It appears that protein is the most influential macronutrient for appetite. Moreover, independent of the modifications of protein or carbohydrate, fat is the least satiating macronutrient, which is in agreement with a couple other studies (here and here) that showed increased hunger after consuming a high-fat meal relative to high-protein and high-carbohydrate meals. Ultimately this led to some great advice by the study authors,

“Although there were no differences in subsequent energy intake, adjusting the nutritional profile of a meal, especially replacing fat with protein, could make dieting more endurable.”

Interestingly, there were no differences in appetite sensations between most of the test meals regardless of protein content, suggesting that the protein and carbohydrate content of a meal may act synergistically. This may help explain why a high-carbohydrate, low-protein food such as potatoes are one of the most satiating foods known to us.

The differences in appetite responses observed after the intake of the various test beverages appeared to be modest and were not followed by differences in energy intake at the subsequent lunch. As already mentioned, the sex differences in energy intake could possibly explain why. However, it could also be that liquid calories are not as satiating as solid food.

Although not a primary outcome of the current study, the researchers did also investigate differences in the metabolic and hormonal responses to the test meals. As could be reasonably expected, postprandial glucose and insulin profiles reflected the carbohydrate content, with increased carbohydrate resulting in the greatest insulin and glucose responses.

However, the HP/HC induced the highest insulin response, whereas the HP/LC induced one of the lowest insulin responses compared with the other test beverages. This highlights the synergistic impact of carbohydrate on the effect of protein on insulin. This effect could be mediated by mechanisms related to increased concentration of amino acids, and there is some cool long-term data showing that a high-protein diet increases pancreatic beta-cell sensitivity to glucose.

Unfortunately, the current study does have limitations. For one, the study population was young, normal-weight, and healthy. For two, there is evidence that the type and source of macronutrients may influence satiety, appetite-regulating hormones, and energy intake differently. In the current study, the protein was from milk protein isolate, the fat from rapeseed oil, and the carbohydrates from maltodextrin and sugar. How this would compare to steak vs butter vs potatoes remains to be determined.