Thursday, December 20, 2012

“We talk about stress all the time”: A case study of stressed and less stressed days among a small sample of undergraduates at a private university




I will admit, I borrowed the first part of the title from one of the student participants in the data you are about to see.  As part of my upper level “Research in Human Biology” class, students collected saliva samples across two days – one self defined “stressful” day and one “less stressed” day. I deliberately did not define stress or less stressed days. This was, in my own way, a means of measuring their subjective experiences of stress.  While the design of this class project came out well before the US News report, the timing could not have been better. My institution was ranked in the top 5 for undergraduate stress using a clears throat – flawed metric. It did provide however, a jumping off point for both this and the report we will have in a month.  Stress was something students were already discussing – albeit a subjective measure. What the in class data collection promised to do was provide an objective, physiological measure of stress by measuring the amount of cortisol, a hormone involved in the stress response, in saliva samples.

Ideally, when measuring salivary cortisol, one would collect between 3 and 5 samples per day (Adam et al., 2010). However, in an effort to simply the collection protocols, we collected two samples – one at waking and one at bedtime.  Again, samples were collected on two self defined days – stressed and less stressed. I had intended a “not stressed” day, but given current levels of student involvement in extracurricular activities, heavy course loads, and high expectations associated with each course, this became more of a fantasy on my part. I think faculty want to think that students exist only in an academic sphere – dorms are places of sleep, study, and deep discussion about how awesome their classes are, merely a bridge between academic endeavors. Having now collected 48 hours of activity data from these same students, I can assure you, they are busy. Tremendously, mind-boggling busy. It is quite impressive really.

The average waking cortisol sample for the class was 0.368 (0.166) ug/dL and the bedtime/pm sample mean was 0.274 (0.160) ug/mL.  For self reported “less stressed” days, morning cortisol was 0.375 (0.177) and evening cortisol mean was 0.374 (0.177). Stressed days had an AM mean of 0.382 (0.22) ug/dL and an evening cortisol of 0.165 (0.146) ug/dL. To see that graphically, check out Figure 1, below.

Figure 1

So how do we interpret these findings? First off, I am not a stress researcher . . . I just play one in the classroom. My own passion, as the title of the blog suggests, is human milk. But, I did do some digging.  What do these cortisol values mean (if you are looking at references with nmol/L, the conversion factor is the ug/dL times 27.59).  Aardal and Holm (1995) give reference values of 14.6 nmol/L (0.56 ug/mL) as the mean for similarly aged females and 10.0 nmol/L for males in the morning (waking) and 1.1 nmol/L for females and 2.2 nmol/L for males in the evening (bedtime).  If there are more recent articles with clearly defined stress cut-offs, I am happy to revise this post – but I did not find any on PubMed and calls via Twitter went unanswered.  Compared to the Aardal and Holm reference values, physiologically, the students in my class are modestly but not severely stressed.  Waking cortisol was 10.16 (4.57) nmol/L and bedtime was 4.97 (4.06) nmol/L (see above for values in ug/mL).  Waking cortisol was slightly lower than average, whereas bedtime cortisol was high compared to the reference standards. This suggests that the students may have a blunted cortisol response. 

Figure 2

This was true, and in a few students, salivary cortisol levels actually increased over the day (Figure 2) compared to the physiologically normal/ideal cortisol pattern of a decline from waking to evening (although the relationship is not quite that simple). Evening cortisol levels on stress days were often much lower than evening cortisol levels on less stressed days. The students chalked this up to “the stress being over” by bed, as their stress days were typically defined by exams, presentations, or interviews that were finished well before bedtime.  Figure 3 shows a comparison of morning and waking cortisol levels by individual on stressed and less stressed days (colors are the same individual). 

  Figure 3a: Less stressed

Figure 3b: Stressed

Wednesday, October 31, 2012

Moving forward . . .

Last night, despite the grants that needed to be revised and the classes that needed to be prepped (woe is me), I managed to do a few extra pages in my book. I usually manage a handful of pages a few times a week. Normally, this reading is not really worthy of mention - either I am reading something vaguely related to research or a novel. Lately however, I have been reading "Don't be such a Scientist" by Randy Olson. I think I paid $2 for it; it has increased in price.

Anyway, it has been an interesting read. Looking over the reviews on Amazon, most people found it reasonably helpful; a few did not, and a few reviews are a little out there. However, I was thinking about offering to run a reading group on it with graduate students. Olson makes some great points about how science communication contributes to the lack of science knowledge, and may also be an important factor in the push back against science and education. He argues that in communicating science, scientists often come off as arrogant or unapproachable and sometimes down right mean.

I am not going to argue that any time I have to deal with creationism I am completely level headed. I fume. I huff. I draft a response and redraft, revise and pace. In fact, I dealt with such an issue earlier this week. And finally broken down and confronted the individual about the material, framing it from a logical stand point - your references are too old, why is up with that? Confusion poured out. In this case, creationism was confusion. In some ways (this was a student) I had failed. I had been teaching them concepts for 8 weeks, and the student could not apply the knowledge from class to the website they had found the material. We had a good chat. We decided to revise the post using more recent research (from the last 20 years) and will ultimately talk about how to tell peer reviewed writing from a website and evaluate sources.

This, and what is going on here (see the prior post) is what prompted me to really start thinking about how I am communicating science. The second point Olson makes (of many, but that are important to my current thinking and worldview - not to mention I am only 63% into the book) is about science blogging. Specifically, how much anger and frustration is poured out into science blogging. Olson suggests that this anger is the first motivation to put key to pixel, and most blogs start with this early reaction against something. Good blogs either do not start this way or quickly move beyond this. I have now committed the first cardinal sin of blogging, and I plan to move forward in the level headed way. I also plan to get students from my lab group blogging here as well. Hopefully, this will keep the conversation from being too focused on the stuff I find cool and will keep me from sounding too much like a scientist.

So moving forward, we will be posting. We will be learning how to use shiny new features, like pictures (I know, not new or shiny). And we will be thinking. And communicating. And not letting anger or frustration drive posts but a joy for the subject.

Thanks for reading!

Monday, October 22, 2012

On undergraduate research, active learning, and IRBs

Today marks another day in the long, slow process of obtaining IRB approval for my class project. It is moments like these, when after a month of back and forths, I begin to really think about Human Subject boards and what they mean for research. First and foremost, I am a huge fan of Human Subject approval boards. I think the history of human based research provides plenty of bone chilling examples why we need IRB boards.

However, I am less than convinced that IRB boards completely fulfill their stated goals. In many instances, the behavior of boards would seem to be directly contradictory to their mission. This summer, I received a mail based survey, approved by an IRB and sent to my home that requested my social security number. Let's take a moment and look at this particular example. The IRB board was, in this (as in all instances) supposed to be protecting my privacy and rights as a research subject. However, they approved a study that 1) targeted people living in my zipcode, 2) had my address, and 3) wanted my social security number? The steps above are a how to guide for identity theft!

I, as I am usually on the opposite side of the IRB process and had just successfully had the requirement removed from my own IRB, was understandably vivid. There were a few choice calls made.

Which brings me to today's IRB challenge. Or active learning is a headache . . .
My research methods class designed their own survey, and wants to test the survey with their peers. So, as we all know from Bloom's taxonomy and the resulting body of research that followed, one of the best ways to facilitate student mastery of concepts is to allow students to design and execute their own projects.  My class did this. They designed their own study, and wrote an amazing survey. They all took citi certification and we discussed research ethics. We even won a research grant to conduct the research.
And, even though it is a class project and might be exempt from the IRB requirements, I submitted an IRB. I want them to be able to actually do something with their data. We asked for exempt status from written consent, as we would not be collecting names.

We found out today we will not be getting the exemption. IRB, in "protecting" subjects (undergraduate students studied by other undergraduates) feels that the best way to protect the students' privacy is to require us to collect their names on written consent forms. Once we have consent forms, the data cannot go home with students but must go straight to a research facility - so by the way would I mind giving them access to my lab 24/7.

I am certain we will come to a solution. And I am also certain as IRB frustrations go, this is on the small end. I think however, it illustrates the true problem with current pedagogy - while the university will sing the praises of active learning, problem based learning, and "undergraduate research" the infrastructure necessary to conduct projects - and to protect students participating in the research - is lacking. Psychology departments once (and many still do) treat participation in faculty research projects as a part of course requirements. There is certainly a power difference there (it is part of your grade to allow us to experiment on you) that we do not have here - this project is student designed, student executed, students interviewing students.

My commitment to this kind of active learning - facilitating the class in designing and executing a project - remains strong. I simply hope that all aspects of the university can better synchronize how we do student designed and student lead research projects in the future.

Wednesday, May 30, 2012

Day 1 - Figuring this out.

First post! And I have no idea how any of this may look . . .