Cortisol Circadian Rhythms and ow Cortisol Concentrations are Affected by Gender, Steroids, and Stress. Michael P. Tekin, and Chandra D. Dewar Department of Chemistry, Bloomsburg University, Bloomsburg, PA 17815 ABSTRACT: In humans, cortisol levels should be highest in the morning and lowest at night. Cortisol levels should also be higher in individual under stress (1), ones that take steroids (3), and males compared to females (4). An ELISA was run on saliva samples taken from thirteen subjects. An anonymous survey was used to gather information about the subjects. The data collected was analyzed using a two-tailed student T test. The cortisol concentrations of the samples were significantly different when comparing samples collected in the morning and night (Difference = 92.20%, T=0.997, p=0.005) but not significantly different when comparing samples collected in the afternoon and morning (Difference=49.23%,T=0.937, p=0.071). Samples collected from males did not have significantly higher cortisol concentrations than samples collected from females (Difference = 42.15%, T=0.937, p=0.080). Steroid use caused a significant decrease in cortisol levels (Difference = 53.76%, T=0.944, p=0.028). Stress levels did not follow the expected trend (1). The only significant difference was that individuals that reported a stress level of 0 and a stress level of 1. Therefore, the data supports that cortisol is regulated via circadian rhythms, but does not support that fully support that stress, steroid use and gender affect cortisol levels. INTRODUCTION Cortisol is a hormone released from the adrenal gland that regulates several bodily functions. Cortisol stimulates gluconeogenisis, glycogenolysis, the breakdown of proteins in the skeletal muscle, and lipolysis; all effects that result in an increase of blood glucose (1). Other effects of cortisol includes the suppression of immune function, the increase in calcium ions (Ca 2+ ) and alterations in brain function. O O C 3 C 3 Figure 1: The molecular structure of the cortisol hormone (2). O O O Cortisol (Figure 1) is released in response to physical and physiological stress (4). Additionally, cortisol is released under normal conditions throughout the day where levels are highest in the morning then pursues to gradually decrease into the night (1). Cortisol levels are also affected by steroid use. If the steroid used resembles or is a derivative of cortisol, measured cortisol levels will be inflated (2). Another variable is gender. On average, when under stress, males release more cortisol than females (4). Because of the volume of variables that have an effect on the cortisol concentration in the human body, certain hypotheses were tested to verify the validity of the aforementioned affects by analyzing saliva samples of several test subjects. The measured cortisol concentrations should, therefore, follow the expected circadian rhythms in the samples collected from the
individuals studied, physical and psychological stress would cause a rise in cortisol measurements (1) and steroid use will inflate the levels of cortisol in the samples tested (3). To confirm the predictions previously stated thirteen Bloomsburg University Biochemistry Students were asked to obtain six samples each of saliva over a specified period of time that would eventually be analyzed to measure the cortisol levels within each sample. An ELISA was run on the collected samples and using a standard curve the concentrations were calculated and analyzed. EXPERIMENTAL PROCEDURES The procedure published by Pugh, et al. (2) was followed. Deviations from standard procedure are summarized as follows: Cortisol collection: The procedures states six samples should be collected over two days; once in the morning, once at noon, and once at night each day. owever, most of the participants collected six samples over three days. The samples were collected by each participant in the morning, afternoon and night beginning at night on the first day and ending in the afternoon on the third day. Data analysis: Standard procedures dictate the background absorbance should be removed from the data when constructing a standard cure. owever, the instrument used removed background absorbance automatically. The cortisol concentration of each sample was compared by the following variables: time of day, steroid use, and stress levels. Stress levels and steroid use was obtained anonymously via an online survey. The Likert Scale for stress levels included the selections of strongly disagree, disagree, neutral, agree, and strongly agree. In the order stated previously the responses were given the following values for statistical analysis: -2,-1,0,1,2. An unpaired Twotailed student T test was used to determine statistical significance. RESULTS AND DISCUSSION To calculate the concentration of cortisol in the unknown samples a standard curve was produced (Figure 2). The standard curve compared %B/Bo and cortisol concentration, in ng/ml, resulting in an exponential equation that best fit the data. %B/B o 100.00 80.00 60.00 40.00 20.00 y = 22.242x -0.449 R² = 0.9883 0.00 0.0 5.0 10.0 15.0 Cortisol Concentration (ng/ml) Figure 2. The standard curve relationship between %B/B o and cortisol concentration demonstrated exponential decay. The precision of the standard curve was indicated by the R² = 0.9883. The equation y = 22.242x -0.449 was used to determine cortisol levels in all unknown samples. The trend of the standard curve displayed an exponential decay achieved from removing the 0.00 ng/ml cortisol data point from the dataset used to produce the curve. The R 2 of the curve indicated the equation was precise (R² = 0.9883). The resulting equation (Equation 1) was then
used to calculate unknown cortisol concentrations. Y=22.242x -0.449 (Equation 1) The calculated cortisol levels were tabulated and equated to display a demonstration of how cortisol levels compare between males and females and individuals whom do or do not use steroids in relation to the time of day at which the samples were taken (Table 1). Averages were taken of the cortisol concentrations of each group being analyzed. Table 1: Compiled results of cortisol concentrations (ng/ml) relative to the gender, steroid use and time of sample collection of 13 test subjects. Time Male Female All Steroid None Morning 5.10 2.32 3.71 2.33 5.09 Afternoon 2.13 2.41 2.24 1.95 5.09 Night 1.67 1.07 1.37 1.07 1.66 All 2.97 1.93 2.44 1.78 3.10 The results of one test subject, subject 23, were removed from the calculations. Individual 23 displayed abnormally high cortisol levels compared to every other individual of the group. Performing a Grubbs Tests to assess the validity of treating the subject as an outlier was performed using an average of the total concentration of cortisol collected from all six samples of each subject. The test performed suggested that subject 23 s cortisol levels were an outlier in comparison to the rest of the individuals tested with a 95% confidence level being certain that there was a probability of less than 5.00% that the decision to remove the data was false. The excessive cortisol levels may be explained by the fact that subject 23 uses steroid inhalers (3). As a result, the cortisol would have originated from an inhaler dosage rather than the body s natural production of cortisol resulting in a dramatic increase in cortisol levels. owever, there are several factors that have the potential to increase cortisol concentrations in the body but relative to the average levels amongst the group suggested a factor, other than gender and time of day, had played a part in the high concentrations and therefore falsely skewed the data when comparing gender and time of day effects. After removing the outlier data, trends were observed with cortisol levels in relation to gender and time of day. After averaging the results of total cortisol produced by the males and females throughout the sample collection timeframe males possessed 42.15% higher cortisol concentrations than females. The difference in cortisol levels was not unexpected since males produce more cortisol when stressed than females (4). owever, the trend was not significant (T=0.937, p=0.080). The small sample size and variable data can account for the insignificance of the results. Also, after averaging cortisol concentration of all participants and partitioning the data to represent the total cortisol collected from the entire group of participants at the three different timeframes throughout the day the quantity of cortisol was highest for samples collected in the morning and lowest for samples collected at night. The morning samples collected and average of 49.23% more cortisol than the afternoon samples and 92.20% more cortisol than the night samples, results that were to be expected. The difference in the cortisol concentration observed among samples taken in the afternoon and morning was not significant (T=0.937, p=0.071). Again, the trend can be explained by the small sample size and variable results. owever, the difference observed among the samples collected in the morning and night was
Cortisol Concentration (ng/ml) significant (T=0.997, p=0.005). The measured cortisol concentration followed the expected circadian rhythms (1). The results obtained from comparing steroid users versus those that do not use steroids produced data that was not to be expected (2). The group of individuals that did not use steroids collected an average level of cortisol that was 53.76% greater than the group that used steroids in some fashion. The difference between the two groups was significant (T=0.944, p=0.028). owever, the figures calculated for the effects of steroid use did not include subject 23 s data. If the reason for 23 s high cortisol levels was the direct result of steroid use then the results obtained from the steroid study are inconclusive. The cortisol levels were also analyzed in comparison to the levels of stress each test subject felt at the time each sample was collected. The subjects were provided with a survey and were allowed to respond to a stress level survey ranging from strongly agreeing to experiencing physical or psychological stresses to strongly disagreeing to experiencing physical or psychological stresses. Each level was then provided a number from -2 to 2 synonymous to strongly disagree to strongly agree. The levels were then combined to obtain data of the overall group, where levels of physical and psychological stresses were added together to result in stress rating levels that ranged from -4 to 4. The resulting data can be seen in Figure 3. 10 8 6 4 2 0-4 -3-2 -1 0 1 2 3 4 Stress Rating Figure 3: Cortisol concentrations (ng/ml) relative to physical and psychological stress ratings experienced by thirteen test subjects. The spotted bars ( ) represent cortisol concentrations that resulted from physical stress, the solid black bars ( ) from psychological stress and the striped ( ) from both stress types combined. The stress ratings represent survey responses that resulted from strongly agreeing (2) to strongly disagreeing (-2) to whether or not a specific stress type was being experienced at the time each sample was taken. Cortisol concentrations tend to be highest when a person is experiencing high levels of stress (1). owever, the observed cortisol levels were lower in individuals that had reported psychological, physical, and combined stress than those that did. Moreover, there were not any significant differences in cortisol levels as stress increase (p>0.05). The only exception was the cortisol levels in samples that reported a psychological stress level of 1 had significantly higher cortisol than and the individuals that reported a psychological stress level of 0 (difference = 282%, T=99407, p=0.0075). The highest averaged cortisol concentration in response to physical stress was when individuals experienced a stress rating of -2 where the highest averaged cortisol concentration in response to psychological stress was when individuals experienced a stress rating of -1. When combining both physical and
psychological stress ratings, cortisol levels were highest at -3. The issue with utilizing the survey to obtain data from each participant originates from the idea that each person had the capability of interpreting each rating differently than another. For example, one individual may rate a stress level as a 2 where another individual may have experienced the same level of stress but rate the experience as a 0. Also, being able to differentiate moderate levels of stress and high levels of stress may be difficult for some subjects. An experiment that induces stress in a controlled setting would be more effective in causing increased cortisol level than procedures used (Lovallo, et al. 4). The sample size of participants was also too small. Thirteen individuals does not allow for a large enough sample set to definitively provide comparable results. Such an issue was observed when analyzing the results of subject 23, the only individual to possess abnormally high stress levels, in comparison to the other test subjects, which was thought to be the result of steroid use. owever, when comparing the result of the other steroid users, steroid use did not appear to have an effect on cortisol levels other than that steroid reduced the levels of cortisol. The survey taken fails to ask the frequency at which steroids were used and when and if the steroids were taken at the time the samples were collected which may have a bearing as to why steroid use did not produce the expected results. Also, because the stress data appears to be seemingly insignificant, having a far greater dataset to analyze may have the capability of producing trends that are currently not available. The results partially support that cortisol is regulated by circadian rhythms because cortisol levels decreased as the day progressed there was a significant difference between samples collected in the morning and night, but not between samples collected in the morning and afternoon. Also, the difference of cortisol levels between males and females and most stress level was not significant. Additionally, the results indicate that steroids decrease cortisol levels which contradict previous studies. Therefore, until previously stated issues are resolved, a large portion of the analysis will remain inconclusive. REFERENCES 1. Silverthorn, D. (2013). uman Physiology 6th ed., Pearson Education, USA. 2. Oxford Biomedical Research. (2008). Enzyme Immunoassay for Cortisol. pp 1-5, Oxford Biomedical Research Inc. Oxford MI 3. Pugh, M. E., Bell, T. T., and Borland, M. G. (2014) Cortisol 1: Analysis of Salivary Cortisol in umans, in Advanced Experiments for Biochemistry 2 (Pugh, M. E., Bell, T. T., and Borland, M. G, eds.) Spring 2014 ed., pp 68-69, Bloomsburg University Press, Bloomsburg PA. 4. Lovallo, W. R., Farag, N.., and Vincent, A. S. Use of resting control day in measuring the cortisol response to mental stress: Diurnal patterns, time of day, and gender effect in Psyconeurodocrinology. 35(8): 1253-1258.