Master's in Nutrition Science: What makes the research in nutrition & physical activity so tricky? (Part 3)

Master’s in Nutrition Science: What makes the research in nutrition & physical activity so tricky? (Part 3)

30 December, 2018

Part 3 out of 4 – we are getting there!

Hello there! Wondering about the life of a master nutrition student at KI in November-December? You shall look no further! Here comes the detailed (and hopefully fun!) review of the third course of the Master’s Programme in Nutrition Science called “Diet, physical activity and fitness – assessment and evaluation” that the author of this blog has been actively involved in for the past several weeks.

So, what is all this jazz about?

How do we know what do people actually eat?

Let me begin with telling you the following: there are plenty of ways to access a person’s dietary habits / caloric intakes, but only very few of them are actually 100% reliable. Throughout the course, we worked more closely with 2 of them: 3-day food weight record and self-designed food frequency questionnaire (FFQ). The latter was developed in groups of 4-5, and we were later supposed to use it to measure our own calcium intake. Interestingly, we got to know that the researchers who develop this kind of dietary “assessment tool”, are usually not interested in knowing the food “info” of a defined day, but rather the usual intake of groups and individuals of a particular nutrient or product/groups of products. However, this indirect method of measuring someone’s dietary intake levels is prone to errors, as

“people tend to forget what they ate yesterday, let alone a week or a month ago”.

An example of FFQ from BMJ

A similar “error-like” problem occurs with the 3-day (kinda a necessary minimum!) food weight record. One of the most common problems with such dietary assessment methods is under-reporting: people have the tendency to only report “healthy” foods that they eat, and are not so keen on writing down that “chocolate bar” they had in-between the main meals. Scientifically it’s called self-report bias. Therefore, the important parameters like the energy intake (EI) of some individuals can be under-estimated, and that can, in turn, affect the data on the population level, unless some statistical adjustments will be made (more on the stats part below 😉)

Photo by rawpixel.com on Pexels.com

Moreover, the necessity to weight and write down “every gram of rice and tomato that you put on your plate” can be time-consuming and put pressure on a study subject, especially if he/she already has a stressful life or simply a tight schedule.

Doing a food-weight record can be especially burdensome for meals & dishes with multiple ingredients. Photo by Pixabay on Pexels.com

Likewise, when assessing the physical activity levels of the individuals, or PAL, a researcher must take several things into consideration. Firstly, a significant day-to-day variation in a person’s physical activity levels (just like with the diet) can introduce a substantial error if PAL is only assessed for a single or too few days. Thus, a careful preliminary calculation of the necessary number of days, as well as your study sample (number of participants) is crucial here.

Calculations? SPSS is your new best friend!

If you have never had statistics before, or barely remember “this stuff” from your bachelor’s programme – brace yourself! …. Ok, just kidding 😀 If you simply put enough work into “finding your way” in G-power and SPSS (the main 2 statistical calculation software programs that we used) – you will be fine! At first, we had the lectures where the basics of different statistical tests and concepts (like p-value, variability and confidence intervals) were revised in a “step-by-step” manner (and the lectures were also recorded!), and then we were given more practical hands-on exercises to train our newly obtained STAT-skills, as well as show our outstanding power-calculation abilities (cannot really relate to that though😅) when submitting our mandatory assignments & take-home exam. That one was quite demanding, but most of my classmates managed well nevertheless (which gives me hope to pass it in the future also! :D)

Hope this won’t be the case for anyone in this program 😀

Fitness tests in real life🏃‍♀️

To create our own set of statistical data to work with (yes, we didn’t just get the random numbers – that was the exciting part!), we had to do some FITNESS and PHYSICAL ACTIVITY tests ourselves – all during the scheduled uni time! I mean…doing a 2km UKK walking test in the forest instead of sitting at the lecture – how cool is that?! Moreover, things like Queens step test, SubMax treadmill test, push-up test & grip strength test were performed by each student at the KI Huddinge gym.

As you can see, some Master Nutrition students quite enjoyed this “additional workout” in the form of the mandatory uni fitness test 😀

We also had to wear waist accelerometers for 3 days to be able to calculate our energy expenditure (EE) levels. These wearable motion detectors are very useful for physical activity monitoring and assessment, as they are commercially available and rather inexpensive, as well as provide relatively accurate data on the person’s EE levels in a “real world” living environment. The other physical activity assessment methods, so-called “golden standards”, are the doubly labelled water method (DLW) and indirect calorimetry (measures oxygen uptake & carbon dioxide production). However, these methods are hard to apply on a population scale, so often cheaper methods like accelerometers or pedometers are used (also for the university students like us :D)

A type of ActiGraph waist accelerometer that we were using ourselves. From: https://psychologie.uni-graz.at/en/health-psychology/research/technical-equipment/

Conclusion: dietary intake and physical activity – which method to choose?

As a researcher in the field of nutrition and/or physical activity, you will always face the following question: which assessment method do I choose to answer my research question? Like I described before, there are advantages and disadvantages about each and every one of them. Below please see a nice and detailed chart that depicts some of the most common dietary methods & their main characteristics:

Participant burden and social desirability bias were the ones I encountered myself! Obtained from: programme lecture slides 🙂

One must also ask him/herself some of the following questions:

What data level is sufficient for my study question?
What foods/nutrients /physical activity do I want to measure? (e.g. diet as a whole/specific food groups or physical activity types / Macro- or micronutrients?)
How precise (close to the truth) do I want to be? (usually expressed as a ’percentage from the true value’)
How motivated are the subjects?
How much time and money do I have?