In recent years, researchers have searched for ways to improve on dietary assessment methods. One tool that is widely available for use is smartphone applications. My meal mate is a smartphone application designed to monitor dietary intake. It was developed as a weight loss guide. However, the capacity for My Meal Mate a potential dietary assessment tool was not yet established. The paper by Carter et al. examines My Meal Mate in this capacity, and attempts to validate the tools use as a means for dietary assessment. Fifty volunteers were invited to record seven days worth of intake on My Meal Mate.
During the seven day period, volunteers were randomly telephoned and asked to complete a 24 hour recall. Total energy and macronutrient intake as measured by 24 hour recall was compared to the equivalent day’s record in My Meal Mate. Group means for the two methods were contrasted using paired t-tests. Pearson correlations described how related the two measurements were. Bland-Altman plots were also utilized to help visualize the agreement between the two dietary assessment methods. The first 24-hour recall was compared to its analogous day in My Meal Mate.
Similarly, the second 24hour recall was also assessed against equivalent day in My Meal Mate. For the sake of comparison, the study also examined the average of the two 24-hour recalls against the average of the two corresponding day’s measurements in My Meal Mate. The average of the two-24 hour recalls was also evaluated in comparison to an average of the entire 7 day’s measurements in My Meal Mate. Smartphone applications, like My Meal Mate, that compute daily nutritional intake can certainly be considered novel developments.
Smartphone applications are unique and innovative largely because of their ease of use, convenience, and availability. Their motility and allows users to record dietary intake effortlessly and in real time. Dietary intake applications are most similar to food diaries, in that they rely on the user to self-report intake. As in food diaries, the user documents as they consume. Because it is completed in real time, this tool should be cognitively simple. It’s progressive completion limits recall error.
The idea is that users register foods from a given meal, directly after having eaten them. The application prompts users to indicate portion sizes, and any additional details they can provide. This can includes fat content, brand name, and even preparation details. However, food diaries are typically openended questionnaires, whereby, people are not limited by a food choice restrictions. My Meal Mate, and other similar smartphone applications have a finite number of food items available in their database.
The intent of limiting the foods available for entry is to decrease the expense and difficulty associated with coding food items. Another reason this dietary tool was structured this way is for ease of use. Logically, it does not seem difficult to record everything you eat on a piece of paper. Practically, however, we know this is not true. Writing out dietary intake can be a tedious task. For food records to work as intended, participants are required to record as they go. This requires notetaking ability twenty-four hours a day.
It may, not be feasible to carry pen and paper with you everywhere you go. In today’s society, people rarely go anywhere without their cell phones. Thus, smartphones offer a widely accessible mechanism for documentation. Providing people with a database of foods at their fingertips compels users to record the foods typically forgotten in traditional food records. However, in doing so, the tool restricts individual variability. This tool may not be appropriate for culturally diverse groups of people. Foods commonly eaten may or may not be included in the database.
In this regard, My Meal Mate is more similar to a Food Frequency Questionnaire. FFQs also utilize a preset database of food items. That being said, My Meal Mate quite different from a Food Frequency Questionnaire. Typically, Food Frequency Questionnaires capture past dietary intake, whereby participants indicate how regularly a particular food was consumed over a specified time period. In contrast, My Meal Mate depicts recent diet. Therefore, My Meal Mate is not an adequate means of measuring past diet. In my opinion, My Meal Mate is most diverse from 24-hour recalls.
My Meal Mate does not rely on participants to remember the foods they consumed. As previously mentioned, this tool utilizes on the gradual input of food consumption data. I think that it is interesting that the paper used 24-hour recalls to validate My Meal Mate’s competency as a dietary assessment method. It is my impression that the two methods are too drastically different to make a valid comparison. 24-hour recalls are typically conducted by an interviewer, who knows exactly how to cognitively prompt the respondent.
As technology becomes progressively more important in this ever-changing world, scientists have developed computerized versions of 24-hour recalls. An example of a technologically sawy 24-hour recall is the ASA 24. The ASA attempts to use technology in lieu of an interviewer to stimulate cognitive recall of dietary intake. I think that My Meal Mate is similar to computerized 24-hour recalls in this manner. Both have developed complex computerized programs to encourage accurate reporting, especially of foods commonly overlooked. The study by Carter et al had some very interesting results.
No statistically significantly difference in mean energy intake was detected between the first 24-hour recall and the equivalent day in My Meal Mate (an insignificantly different difference of 68 kJ was observed). No statistically significant differences in macronutrients were noted for first day’s measurements. A statistically significant energy intake difference of 441 kJ was noted between the second 24-hour recall and its corresponding day in My Meal Mate. A statistically significant difference in dietary fat intake was also detected among the second day’s measurements.
Interestingly, the measurements of other macronutrients did not display a similar trend: no statistically significant differences existed among protein or carbohydrates. No statistically significant difference in energy or macronutrients existed when comparing the average of the two 24-hour recalls against the average quantity of the two corresponding days in My Meal Mate. Similarly, no significant differences existed when comparing the average of two 24-hour recalls against the average of the entire 7 day diary recording in My Meal Mate. Correlations between the two methods were remarkably high.
The correlation for the first 24-hour recall was calculated at . 77. For the second 24-hour recall, the correlation coefficient was found to be . 85. A correlation coefficient of. 85 was also found when comparing the average of the two 24-hour recalls and an average of two days measurements on My Meal Mate. Finally, when comparing and average of the two 24-hour recalls with an average of all 7 days on My Meal mate, a correlation coefficient of . 68 was calculated. The Bland-Altman Plot for the first day shows a small mean difference, and the limits of agreement are wide.
Even though the difference is small, a wide limit of agreement means the results may be abstruse. The BlandAltman Plot for the second day has a larger mean difference than the first, and the limits of agreement are slightly smaller. Even though the results have a large difference, their predictability is high. The Bland-Altman Plot for the average of two 24-hour recall and two days of My Meal Mate has the smallest average difference and smallest limits of agreement. Therefore the average of the two 24-hour recall and average of two days of My Meal Mate are most in agreement.
