3.1. Respondent Demographics
The survey was delivered to 1243 students, with 493 respondent surveys being completed and used for data analysis. Participants who did not reply to the survey, did not open the survey, or completed less than 95% of the survey were not considered in the statistical analysis. A total of 453 respondents completed the survey in under 6 min (5.4 ± 3.4 min), with a small number (n = 40) taking longer to submit the completed online survey. Respondents who completed the survey consisted of 105 males and 388 females aged 18.9 ± 2.2 years old, with a BMI of 24.4 ± 4.8, and an ACT score of 23.7 ± 3.1 upon university admission. They consisted of Freshmen, Sophomores, Juniors, and Seniors representing 71.1, 17.9, 7.0, and 3.7% of respondents. Respondent nutrition training in prior classes was classified as either none, 1–3 credit hours, or 4–10 credit hours, representing 52.9, 40.2, and 7.0% of respondents.
3.2. Ability to Write an NNS Definition
Respondent NNS definitions were examined and given one point each for the presence of words related to (1) chemical nature, (2) taste, (3) caloric content, and (4) non-nutritive quality, with a total score of four indicating the presence of all four components and zero indicating none of the components or that the respondent could not provide any definition. The overall NNS definition score was 1.1 ± 1.1. Written definitions that included none, one, two, three, or four of these NNS components were provided by 41.6%, 20.7%, 26.1%, 10.6% and 1.1% of respondents. As such, 62.3% of total respondents were unable to provide any definition or provided one or none of the four target components. NNS definition score was not associated with gender, age, BMI, ACT, completion of prior nutrition classes, or history of weight loss attempt. Freshman respondents had NNS definition scores (1.0 ± 0.1) that were significantly lower than sophomores (1.3 ± 0.1) or seniors (1.5 ± 0.3). Self-reported use of food labels was recorded by 55% of respondents whose NNS definition score was 1.2 ± 1.1 while non-label user NNS definition score was significantly lower (0.9 ± 1.0).
NNS definition score was correlated with the self-estimated frequency of NNS dietary intake (Table 1
). However, data presented later in this study suggests that respondents had difficulty identifying the NNS they claimed to use in this survey. The poorest NNS definition scores were associated with those who did not know if they consumed them, and the best scores were observed for persons who claimed to never consume NNS (1.56 ± 1.04), although this group represented only 5.2% of respondents.
3.3. Ability to Name Examples of NNS from Memory with a Fill-in-the-Blank Format
After being provided with a working survey NNS definition for this study, respondents were asked to demonstrate their depth of NNS knowledge by writing the names of every NNS they knew from memory. A total of 576 potential words were provided, and respondents could name 1.0 ± 1.1 NNS from memory by TN or CN. Of respondents, 37.9% could name no NNS examples, 39.1% could name one NNS, 11.3% could name two NNS, 9.1% named three NNS, 1.5% named four NNS, and 0.4% named more than four NNS examples. With respect to CN, none, one, two, three and four of the NNS examples were identified by 44.0%, 37.9%, 10.3%, 5.9%, and 0.4% of respondents. With respect to TN, none, one, two, three, four, and seven of the NNS examples were identified by 38.1%, 39.7%, 10.3%, 9.9%, 1.0%, and 0.4% of respondents. The number of fill-in-the-blank TN examples (0.9 ± 0.9) was significantly greater than the number of CN examples (0.2 ± 0.6).
The influence of respondent demographics and habits on ability to name examples of NNS by TN and CN in the fill-in-the-blank exercise was also examined. NNS identification by TN was significantly better among females (0.9 ± 0.1) than males (0.6 ± 0.1). Ability to provide TN examples was also significantly influenced by having completed prior nutrition classes, with those having completed 0, 1–3, and 4 or more credits earning TN scores of 0.7 ± 0.8, 0.8 ± 0.9, and 1.2 ± 1.1, respectively. While the correlation between ACT and the number of NNS identified by TN was statistically significant, the Spearman’s correlation was 0.1151, therefore the association was very weak. TN association with BMI, age, and class was not statistically significant. CN identification was significantly associated with class rank. Seniors, Juniors, Sophomores and Freshman identified 1.0 ± 0.2, 0.9 ± 0.2, 0.7 ± 0.1 and 0.4 ± 0.1 NNS by CN, respectively. In contrast to TN, males were significantly better at identifying NNS by CN (0.8 ± 0.1) than females (0.5 ± 0.1). Ability to identify CN examples was also significantly associated with BMI, age, and ACT, however the Spearman’s correlation values were 0.1086, 0.1390, and 0.1365, respectively; therefore, these associations were very weak. Completion of prior nutrition class credit was not associated with a statistically significant ability to identify NNS by CN in the fill-in-the-blank exercise.
Respondents who reported use of food labels provided 1.2 ± 1.2 NNS examples from memory, while those not using food labels provided significantly fewer NNS examples (0.8 ± 0.9). Respondents who had tried to lose weight in the past and who had never tried to lose weight provided 0.9 ± 1.1 and 0.9 ± 0.9 NNS, which was significantly less than those who were currently trying to lose weight (1.2 ± 1.1 NNS). Ability to provide NNS examples from memory was best for those who attempted to never consume them (1.4 ± 0.8) and worst for respondents who did not know if they consume NNS (Table 2
The number of each kind of specific NNS name written into the fill-in-the-blank portion of the survey was also examined. Responses classified as a TN included Splenda, Stevia, Sweet’N’ Low, Truvia, Equal, NutraSweet, and SweetOne, which were identified by 43.8%, 12.8%, 10.3%, 5.9%, 4.9%, 0.6%, and 0.2% of respondents, respectively. Responses classified as a CN included aspartame, sucralose, saccharine, erythritol, xylitol, acesulfame, cyclamate, and neotame; these were identified by 10.1%, 2.8%, 1.6%, 0.8%, 0.8%, 0.6%, 0.2%, and 0.2% of respondents. In the fill-in-the-blank portion of the survey, 6.1% respondents incorrectly identified “corn syrup” or “high fructose corn syrup” as being a NNS.
3.4. NNS Recognition by TN, CN and Decoy Name Using a Click-Drag-Box Format
Respondent ability to use a click-drag-box
to successfully identify NNS (with familiarity prompting where they saw actual survey words) was used to evaluate relative ability to recognize six CN and their respective TN, a set of six decoy (false) NNS names, six common caloric sweeteners, and six food items (Figure 2
). TN were correctly identified as NNS 4.9 ± 1.0 times while CN were correctly identified as NNS 3.9 ± 1.9 times (Figure 3
). Decoy names were incorrectly identified as a being a true NNS 4.7 ± 1.3 times (Figure 3
). The tendency to identify CN as a NNS was lower than the tendency to identify a decoy as being a true NNS. The tendency to incorrectly identify decoy names as being as true NNS was not significantly different from the tendency to correctly identify TN as being an example of a true NNS. Respondent ability to use a click-drag-box
to identify NNS by TN ranged from a high of 95.7% for Sweet ‘N’ Low to 34.5% for Monk fruit (NNS sourced from a raw food were grouped as a TN in this study) with the average being 4.9 ± 1.0. Ability to use a click-drag-box
to identify the NNS by CN ranged from a high of 73.3% for Aspartame to a low of 52.2% for Sucralose, with the average being 2.9 ± 2.2. McNemar’s test was used to show that respondents were significantly better able to identify click-drag-box
NNS by TN relative to each paired CN for all six comparisons.
The decoy NNS names, Aquasweetener, Glucataste, Murina, Neotaste-S, Rubralose, and Xana Light were identified as a NNS by 93.7%, 65.5%, 79.1%, 84.6%, 56.9%, and 89.7% of respondents. The summary score for decoy names (4.7 ± 1.4) was not significantly different from TN, but was significantly greater than CN (Figure 3