Maybe the samples were collected some minutes after the stressing procedure, when the nurses were in some type of "recovery" phase? Did you used salivettes?

How did you measure stress? What cortisol method did you use? Is this the cortisol assay result from a single sample from each person, or did you calculate a slope?

Thank you Jorge and Maryanne,

We used salivettes to collected the samples. Nurses were taught about procedures the day before. All the data were collected at the same hour (7 to 9 am). Only one sample a day for cortisol. After collection of saliva, by one member of the research team, nurses filled an inventory to measure chronic stress related with the work. Results showed moderate levels of stress and 1/4 of the sample with cortisol level below the reference values for age and gender. Nobody had values above. The Spearman correlation (mean of cortisol and mean of stress) showed a negative correlationship (r=-.304, p=.006).

I have teo things in mind:

1- Did you assessed for coping styles with the job of these nurses?

2- You talk about moderate levels of stress... Maybe the stress experienced by the nurses is not as strong as necessary to induce cortisol response...

3- Have you used the quartiles or the terciles of stress ratings to cut the sample in the groups? Maybe in the the higest stress rate group you can observe an effect of stress on cortisol that is "blocked" by the other participants with lowest stress response...

Well... I had three things in mind :D

Jordi

Dear Jorge Ulrich,

thak you for your replay. I had tought about that reasons too. One of the limitations of our study is that we didn't look for psychosocial variables as coping, neuroticism, social support, and so one. Our hipothesis was the relationship between workload, stress, stress symptoms and cortisol. In matter fact we found levels of workload high (mean 46,21h/week) but moderate stressand without correlationship. And for this, those kind of variables were important to understand this absence of relation. But when we divide the sample in three groups of stress levels (low, moderate and high) using the mean and more and less stand deviation, and we look for the relationship between cortisol and stress (using only the moderate and high group) we found this strange result of negative relationship.

My dout is if it is possible that, as we are measuring chronic stress (refered to last 6 months) it could be some kind of desruglation on the circadian production of cortisol...

And, I know, this is another limitation because we don't have a second colection on the day...

But, did you think we could explore this possiblity of less production of morning cortisol for persons under chronic stress with desregulation of the circadian rythms? Note that all of the nurses had sleep in the night before, they didn't work in that night and we didn't found any difference between those who does night shifts or not..

Regards,

Aida

Aida,

This is puzzling. It would be interesting to replicate this study, adding measures of coping and social support. Do you have extreme outliers in the low and high groups? There are some clinical characteristics that can affect results, e.g. endocrine disorders, hormone therapies, pregnancy, breastfeeding. Some researchers prefer to use cortisol slopes, where a flat slope indicates chronic stress. I think you have already thought of these things. I wish I could be of more help! Good luck,

Maryanne

Dear Maryanne,

Thank you for your replay and sugestions. Measures of coping and social support will be important in future research. How many samples a day it is necessary to determine if there is a flat slope? Two samples per day are enough?

Aida

Dear Aida,

There is considerable dispute about how to collect cortisols, how many samples per day, and over how many days. We used the passive drool method, collecting at waking, waking + 30 minutes, and at bedtime (before brushing teeth), over three days. Subjects were advised not to eat or drink within 30 minutes of collection. The three timepoints will allow us to compute the morning rise (slope from waking to waking +30), as well as the area under the curve from waking to 9 PM. Two values are required to calculate a slope. You could also calculate an average cortisol value for the day, or average across days, or calculate areea under the curve, averaged across three days.

Some helpful references in this area might be: (1) Adam EK & Gunnar MR. Relationship functioning and home and work demands predict differences in dirunal cortisol patterns in women. Psychoneuroimmunology 2001; 26: 189-208. (2) Matousek RH, Dobkin PL, Pruessner J. Cortisol as a marker for improvement in mindfuless-based stress reduction. Complementary Therapies in Clinical Practice 2010; 16: 13-19. (3) Hellhammer J, Fries E, Schweisthal OW, Schlotz W, Stone AA, Hagemann D. Several daily measurements are necessary to reliably assess the cortisol rise after awakening: State- and trait components. Psychoneuroendocrinology 2007; 32: 80-86.

Maryanne

were the nurses blind to the study and relevant procedure?

Or were they happy that finally their heavy work load will be demonstrated and "appreciated" scientifically?

Thank you Maryanne for references and good advice to procedures of colecting cortisol.

Nurses were informed about research aim and, of course, they wanted to colaborate and demostrate how stressed they were. But when they filled the inventory the result wasn't the expected... :)

Aida

This situation reminds me of an anecdote from "Robert Sapolsky"s "book why zebras dont get ulcers". he describes a "test" done for a TV documentary where cortisol samples from drivers are collected during the rush hour where traffic is terrible and most stressful for any driver. And guess what the outcome is...

A driver that always goes crazy during traffic had very very low cortisol levels when he was sampled during rush hour. Because he was simply too happy to be chosen, be the person to draw attention to an annoying problem.

Similar to that anecdote. I guess blinding is the problem here. Stressed Nurses were too happy to take part in a stress project!!!! Something out of the ordinary boring tiring routine.

I would assume if you have taken blood samples without informing the nurses of the experiment the results would have been as expected.

You may want to do a comparitor. IF you are using saliva samples, would compare the whole diurnal profile, including nighttime. I would compare workday and night on two consecutive days, then a weekend or vacation night, ie a non-work night. You will have a better reading on your data this way

Some potential reasons include: Were these nurses shift workers which may have upset their diurnal rhythms? It was unclear whether you measured saliva or blood, but presumably if blood then women on oral contraceptive were excluded because of the effect on CBG. I think the most likely explanation though is the HPA axis adaptation to repeated acute stress results in lower cortisol responses to the same stressor. Those who were most stressed on a repeated basis may have down-regulated their axis.

Dear Aida mendes,

You could read this article. It is very informative.

The relation between work-induced neuroendocrine reactivity and recovery, subjective need for recovery, and health status.

Sluiter JK, Frings-Dresen MH, van der Beek AJ, Meijman TF.

J Psychosom Res. 2001 Jan;50(1):29-37.

PMID: 11259798 [PubMed - indexed for MEDLINE]

Regards

Kerstin Wentz