Question 1
Dietary intake data from NHANES 1999-2000 and 2001-2002 was based on one 24-hour dietary recall; intake data from 2003-2004 was collected on 2 days and then averaged. Do you think this is the most accurate method for collecting dietary intake or assessing overall whole grain consumption? How could this part of the study be improved?
posted by Kevin @ 8:34 PM
31 comments
31 Comments:
I do not think this is the most accurate method to collect data, but it is the cheapest and easiest. I think that a 24 hour recall is probably the only way to collect with such a large sample but it would be more beneficial to average more than 2 days. I think it is important to get a couple week days along with a weekend day for best results.
On average, I feel that I consume a fair amount of whole grains and make it a point to chose those types of foods. But when thinking back about what I ate today, I have not consumed any whole grains. This was not a typical day for me. This is an example of why I do not necessarily believe that a 24-hour recall is the most accurate way of measuring dietary intake. I would say that collecting and averaging at least 3 days of intake would be more accurate. Also, I feel that a food frequency questionnaire could have been useful in this situation. This could have given a better overall view of an individual’s usual dietary intake instead of focusing in on one single day.
I do not think that this is the most accurate way to collect data. I think the data collection should have remained consistent throughout the study. Averaging the data from 2 days is the best method because it allows for more food intake to be recorded and may include one's typical day as Brooke mentioned. I think if they had started the study off with a 3 or 4 day average, the data would be more valid although they may not have received as many responses. With the amount of participants in this study, I'm sure it would have been more difficult to get more days out of them.
I agree with Stephanie that it is important to average more than two days if possible including weekend and week days. Doing the data collection this way may have provided more accurate results but would have been more costly.
I do not think that a 24 or 48 hour dietary recall is the most accurate method to collect dietary intake and assess overall whole grain consumption. A smaller study could have been done, not using NHANES data, so that more information could be gathered and used from each participant. However, I'm unsure whether the benefits of gaining more information from a smaller number of participants would outweigh the benefits of using reliable, NHANES information from such a large number of participants.
I wonder why a FFQ wasn't used. I feel a "snapshot" of one day's diet does not capture the entire eating habits of an adolescent. Due to the high volume of the study, a 24 hour recall is the quickest and most cost-effective, like Stephanie had mentioned. Like Brooke said, if you asked me to do a 24 hour recall of a Saturday...there may not be any whole grains in there =)
I agree with all of you when you say that averaging two 24 hour diet recalls is not the most accurate method to collect data on whole grain consumption. The most accurate form of a diet recall or food record would be a written record of the amounts of food consumed during a time period of at least three days and a maximum of seven. Usually respondents will become fatigued only after several days of recording their food intake. Also, it may become more challenging for individuals of low socioeconomic status, the poorly educated and children if they are required to respond on paper or through the internet. While the averaging of two 24 hour diet recalls is not necessarily the most efficient option, fortunately the advantages and disadvantages equal each other out. Whereas with a food record there are many disadvantages that primarily affect the data being collected and its accuracy. Some main disadvantages of food records or diet recalls consisting of multiple days is that it is a burden on the respondent, it affects eating behaviors, intake is often underreported and it may lead to substantial sample bias. In this case, I believe that while averaging a two day diet recall seems slightly unsubstantial it may be the most cost efficient and respondent friendly to this particular age group.
I don't believe that this is the most accurate way to measure whole grain consumption in this particular study. I feel like they should have only used data from 1999-2000 or 2001-2002. Having the mix and match approach makes the method unreliable to pull results from. However, I do think that recalls are the only way to conduct a large study like NHANES. I believe this because the whole nation is not going to be able to fill out food diaries. If they did it would be full of errors and people curtailing their diets to what the researchers want to hear. So overall, I believe the recall method is best for this study but they need to be consistent on the amount of days they use either 1 or 2 days.
I agree with what many of you have said on here. Brooke, Stephanie, Amanda I feel like they need to take at least 2 or 3 days and average them. They also need to include a weekend day in there. I don't feel like it would be that much more costly and would increase the accuracy of the results to a degree where we could see much more valid correlations between who is consuming whole grains regularly and their health condition.
This is not the best assessment strategy to determine accurate whole grain consumption because one or two days is not indicative of overall diet. A food frequency questionnaire would have been a better tool to use in my opinion for assessment.
Emily, I found your comment that using a smaller study with more information from each participant may not be as reliable as using NHANES data very interesting. My first thought when reading this study was that another smaller study should have been done to see hwo accurate the whole grain consumption from the NHANES data compared to a smaller representative population really was. This would give us insight on how accurate the 24 hour recalls were in measuring whole grains.
I do not think this is most accurate because their may be many factors that may affect any given day or two in an individuals life. Thus I believe a Food Freguency questionnaire may be more appropriate to determine the actual amount of whole grains these individuals are taking in rather than assuming one day represents any given day.
I agree that this was not the most accurate way to assess intakes but it was probably the most feasible. As we've seen with other studies, size really limits the methods researchers are able to use. I think a food frequency questionaire or a 3 day food record would have improved the accuracy of the results.
Emily and Michela, I think the idea of conducting a smaller study with more reliable methods is interesting. I think it would be beneficial to compare the results from a smaller study conducted with a food frequency questionaire to the results from the larger study with diet recalls. I wonder how similar, or different, they would be!
I think with a study this large it is the only way to collet data. It would not be possible to collect and analyze the data with out a large budget and staff. The study would have stronger results if they collected data over a longer period of time. Only evaluating 3 days over a span of 12-19 years does not seem sufficient enough to draw any firm conclusions based on this study.
Jamie you make a good point that averaging the two days is better than only viewing one. I think that 3-4 would have strengthened the results. Although the participation may decrease I think it would be worth it to have more reliable data.
Molly, I agree that the food journal has the potential to be very accurate, but for a study of such a large size it's probably not the best choice. It also requires a lot of work, motivation, and accuracy by the participant. After reading everyone's comments I am even more convinced that the FFQ would have been better method than the 24-hour recall due to its ability to look at a trend in food consumption instead of just one or two days.
I don't really think this was the most accurate method to collect intake. I know from my personal life that some days I eat much better than others depending on the amount of time I put into preparing meals etc. I think the best way to determine intake is possibly averaging at least three days or asking what they typically eat.
Stephanie-
I like your idea about adding a weekend day to the data. Most people's diet completely changes over the weekend. Good point.
Many of you brought up the point of using a FFQ instead of a 24hr recall. I also agree that using a FFQ might have been a better option to collect the data. There would still be possible errors in the collection with individuals over or underestimating their intake. However, every method would have some limitation.
There is really two parts to this question, whether the researchers of the article should have used data that was not consistently obtained and whether a 24 hour and 48 hour diet recall is sufficient enough to accurately describe someone’s diet. The researchers probably should not have used the data from 2001-2002 because it was a 24 hour recall compared to the 2003-2004 48 hour diet recall and a 24-recall does not give a very accurate representation of someone’s diet. I do not believe this was the most accurate method for data collection but because they obtained data from NHANES dating back to 1999-2004 they didn’t have any other options. Ideally I think they should have used a food frequency questionnaire and I don’t see why it would be any more difficult than a diet recall.
I agree that a 24-hour recall is not the most accurate way to assess whole grain consumption, especially when a 2 day record has been used previously. I think that the best way to see whole grain consumption would be to get a 3 day food record so that you could request 2 weekdays and 1 weekend day since most people eat differently on the weekends than during the week. Then, I would average those 3 days to come up with a more accurate assessment overall.
I do not believe that a 24-hour recall is the best option for collecting data. I agree with Brooke that 24-hour recalls do not always represent what an individual normally eats. The individual completing the recall may normally eat half of their servings of grains as whole grains, but that day they did not. On the other side, individuals who normally do not eat whole grains may have had a few servings that day coincidentally. I think the best tool to use would be a FFQ. If a 24-hour recall is mandatory, it may be helpful to use a FFQ in addition.
I do not think this is the most accurate method for collecting dietary intake of assessing overall whole grain consumption because it is based on a very limited window of what each individual consumes. There are various factors that could contribute to a participant’s high or low whole grain consumption of just one day (e.g., stress, day of the week, etc). Like others have said, I think the recall should be based on food consumption of at least 3-7 days for best results.
I agree with Nate, by combining the NHANES 1999-2000 and 2001-2002 data based on 24 food recall with the 2003-2004 data of a 2 day recall makes the study somewhat confusing. The study should use the same time measurement of food intake, but also based on the large amount of NHANES data the study should maintain using the method of a diet recall.
I am also curious why the NHANES Food Frequency Questionnaire (the gold standard as far as FFQs go) was not used. Was it not administered to these participants and therefore the data was not available, or did this group of researchers choose not to use the FFQ data? I feel that the FFQ would prove more appropriate for the nature of this study (grouping whole grain intake into "none", "low", and "high"). As most of you have already said, a 24-hour recall is not a reliable instrument for extrapolating results to overall dietary intake.
As most of you have concluded, I don't feel that the 24 hour recall is the best way to collect data on whole grain consumption. While some people are creatures of habit (like Amy's husband), others like variety (like my husband who does not like leftovers). The FFQ would be the most feasible option to collect data from a large sample and would provide a more accurate view of a participant's diet and whole grain intake.
Like Nate, I also question the combining of data collected using different methods. I also question the age of the data used. Surely there is data more recent than 1999-2004 available for analysis.
Emily, I liked your idea about conducting a smaller study to see how accurate the data is from NHANES, especially since the data was from 1999 – 2004 and the data collection was not consistent. Like many of you said, a FFQ, would produce more reliable results and I think participation wouldn’t decline much, if any.
Nate and Kelsey, you bring up an interesting point about the data being combine between years. This is confusing and I don' t think it helps the accuracy of the results
I agree that a FFQ would be effective. I think the key to using a FFQ would be to modify one of the questions based on information from a 24-hour recall. So what they would need specifically for this study for instance would be "Do you eat whole grain breads, cereals, etc? How often? How much?"
Post a Comment
<< Home