No differences in afternoon appetite control, eating initiation, or subsequent food intake were observed when comparing two commonly consumed, relatively small (160 kcal) afternoon yogurt snacks, varying in protein content (5 g vs. 14 g protein). In a previous snack study comparing low protein (~5 g protein) vs. high protein (~20 g protein) snacks of 20–80 kcal also found no differences in post-snack hunger, desire to eat, prospective food consumption, or fullness
. The lack of differences observed in this study as well as the current study might be due to either the small energy content of the snacks consumed or the relatively small difference in protein content between the low vs. high protein snacks (5-15 g protein differential between snacks)
. In another study that incorporated a larger snack (i.e. 287 kcal) with similar protein quantities (3 vs. 10 g) as the previously discussed and current study, consumption of the snack with 10 g protein resulted in decreased hunger and increased fullness up to 60 min post-snack. This study also observed a lack of difference in time to dinner request between the two snacks
. An additional study that incorporated larger snacks (i.e. 240 kcal) with greater protein quantities (26–46 g protein), substantial increases in hunger and reductions in fullness were observed compared to normal protein versions
. Because a larger snack size (i.e. energy content) typically occurs simultaneously with larger protein quantities, it is unclear as to which factor has the greatest impact on appetite control and satiety.
In addition to perceived sensations of fullness over time, another indicator of satiety includes the onset of subsequent eating (i.e., eating initiation). Most of the current research utilizes a fixed meal design resulting in the consumption of a subsequent meal without taking into consideration the motivational state of the individual (i.e., whether the individual is sufficiently hungry enough to want to eat again and/or would have requested dinner at that time)
[17–19]. Only a limited number of studies, including the current study, incorporate this approach by identifying the time to voluntary dinner request
[16, 25, 27, 28]. Of these, only a subset were snack studies
[16, 27], and only one study previously evaluated the effects of normal vs. high protein snacks
. As shown in Marmonier, et al.
, a high protein afternoon snack led to a 30 min delay in voluntary request for dinner compared to consuming high carbohydrate or high fat snacks
. The lack of difference in the onset of subsequent eating in the current study might again be attributed to the smaller absolute protein quantities (5 vs. 14 g) or the smaller protein differential (9 g) between our snacks compared to the Marmonier et al. study which compared 26 g vs. 46 g protein with a 20 g differential. Therefore, dose-dependent studies of varying protein quantities in varying snack sizes are warranted.
Lastly, by allowing the timing of dinner to fluctuate, we were able to determine that perceived hunger is the best predictor of voluntary eating initiation. Further, in healthy women, perceived hunger reaching a value of approximately 80 mm was sufficient to elicit the onset of eating. Thus, it is now possible to incorporate this data into fixed meal designs based on a specific hunger threshold.
In the current study, several factors exist which might have influenced the overall study findings. Dietary restraint was not specifically assessed, or used as part of the screening criteria. However, individuals who were clinically diagnosed with an eating disorder, those that displayed rapid weight gain and/or loss over a short period of time (≤ 6 months), those that had a-typical eating behaviors and/or patterns (1 meal/day; 6 meals/day, etc.), or those who infrequently or never snacked were excluded from the study. We sought to include a sample size that would be representative of a healthy female population even though it might have included restrained eaters. Menstrual cycle phase was also not controlled. Currently there is conflicting and limited data as to the extent that menstrual cycle phase influences acute appetite control and food intake
[29–32]. Recent data from our laboratory suggests that there is no effect of menstrual cycle on these outcomes
. However, by randomizing the order of the snack conditions between subjects, it is likely that we had an equal distribution of testing days that fell in the follicular and/or luteal menstrual phases. The relatively small energy protein differential of these snacks may not be sufficient enough to elicit protein-related improvements in appetite control, delays in eating initiation, and subsequent energy intake. Lastly, we recognize that these results are specific to healthy premenopausal women and cannot be extrapolated to other populations. Thus, further research to determine a “protein threshold” to elicit the proposed protein-related improvements in a more diverse population is warranted.