Serum and Salivary Cortisol Levels in Diabetics Attending University of Calabar Teaching Hospital, Calabar, Nigeria

Aims: To investigate if salivary cortisol can be used as an alternative to serum cortisol in the management of diabetes mellitus. Study Design: The design of the study was cross sectional. Methodology: Fifty five (55) type II diabetic patients and thirty three (33) non-diabetic controls (45 men, 43 women; age range 30 - 69 years) were recruited for this study. The levels of salivary cortisol, serum cortisol, fasting plasma glucose and glycated haemoglobin were determined. Serum and salivary cortisol was determined using enzyme immunoassay; fasting plasma glucose using colorimetric method and glycated haemoglobin using cation-exchange resin separation method. Results: The mean serum and salivary cortisol levels were significantly higher (p=0.000) in diabetics as compared to the controls. The salivary cortisol was about 70% lower than the serum cortisol in each group. There was no significant difference between the percentage difference between the serum and salivary cortisol levels in both the diabetics and controls. There was a significant positive correlation between serum cortisol and salivary cortisol in both the diabetic subjects (r = 0.362, p=0.007) and controls (r =0.406, p= 0.019). Conclusion: Cortisol levels in saliva reflected those in serum in both diabetics and controls therefore salivary cortisol may be used in place of serum cortisol in the management of diabetes.


INTRODUCTION
A high prevalence of subclinical hypercortisolism has been suggested in patients with type II diabetes mellitus with poor metabolic control and several observations have suggested that in type II diabetes patients, subclinical hypercortisolism may be more frequent than previously expected [1]. Elevated levels of cortisol have been reported in type 2 diabetics [2].
The important role cortisol may play in the development of type 2 diabetes has renewed interest in adrenal function in diabetic conditions. It is probable that small increases in cortisol levels, even within the range of normal, may cause a deteriorating effect on diabetes itself, thereby increasing the risk of diabetes-related complications [3]. Diabetes (mainly type 2 diabetes) has been shown to be associated with cortisol levels because hypothalamic-pituitaryadrenal activity is enhanced in patients with diabetes complications and the degree of cortisol secretion has been shown to be related to the presence and number of diabetes complications [4]. High levels of cortisol in the body increase blood glucose levels through glycogenolysis and also increase mobilization and breakdown of blood lipids. This leads to hyperglycaemia and hyperlipidaemia. Hyperlipidaemia contributes to insulin resistance. Hyperglycemia and hyperlipidaemia are classic symptoms of diabetes mellitus. An elevated level of cortisol also antagonizes the effect of insulin on blood glucose [5].
Studies have shown that the compliance to tests and therefore good management of diabetes is reduced due to anxiety, fear and weariness of diabetic patients over repeated blood sampling for tests [6]. Anxiety has been linked to worse glycaemic control in diabetics. Therefore, less invasive sampling methods may reduce anxiety and indirectly help in achieving improved diabetes control.
It has been shown that salivary cortisol measurement can be an alternative to serum cortisol as a marker for adrenocortical function following insulin tolerance test, corticotrophinreleasing-hormone stimulation and adrenocorticotrophin hormone stimulation [7]. Salivary cortisol testing is non-invasive, easy to perform, accurate, convenient and cost effective. Salivary steroid levels can reflect the circulating level of free steroid rather than total circulating levels, which are confounded by the presence of circulating high affinity binding proteins [8,9]. It offers the attractive option of a stress-free and real-time repeated sampling where blood collection is either difficult or undesirable [6]. The aim of this study therefore was to determine if salivary cortisol can be used as an alternative to serum cortisol in the management of diabetes mellitus.

Subject Selection
Eighty-eight subjects of Nigerian origin were selected in this study. Fifty-five (55) of these were confirmed type II diabetic patients (males and females) attending the University of Calabar Teaching Hospital (UCTH) Diabetic clinic. They were recruited as test subjects. Thirty-three apparently healthy non-diabetic subjects from Calabar metropolis were recruited as controls.
This was a cross sectional study and consecutive sampling method was used. Ethical approval was obtained from the ethical committee of the University of Calabar Teaching Hospital. The purpose and nature of the research was explained to the participants and informed consent was obtained from them.

Inclusion criteria
All the subjects recruited for the study were 30years of age and above.

Exclusion criteria
Type 1 diabetics, terminally ill type 2 diabetics, as well as anyone who did not consent to participating in the study were excluded from the study.

Sample size calculation
The number of samples used in this research was determined using the formular below After calculating the sample size, a response rate of 80% was assumed therefore the actual sample size was 33/0.8 = 41 cases.

Sample Collection
Fasting blood samples were collected between 8.00am and 9.00am owing to diurnal variation of cortisol secretion. With minimal constriction and stasis, 6 milliliters of venous blood was aseptically collected by venepuncture from each subject. Two milliliters of blood from each subject was dispensed into a tri-potassium ethylene diamine tetra-acetic acid (EDTA) bottle for glycated hemoglobin estimation and 2 ml into a fluoride oxalate bottle for glucose estimation. The remaining 2 milliliters of blood was collected into plain containers for serum extraction, which was used for cortisol assay. The saliva samples were collected into sterile wide mouth containers between 7.30 -9.00 am after an overnight fast prior to the day of the collection. The subjects were instructed to rinse their mouth with clean water severally and discard the water. Then the subjects spat in the container several times until about 1 ml of saliva was obtained. It was then spun at 3000 rpm for 5 minutes and supernatant obtained. Serum and saliva not used immediately was kept frozen till used.

Glucose Analysis
Fasting plasma glucose was estimated using a test kit of glucose-oxidase-peroxidase method produced by Giesse Diagnostics Inc., Italy. Normal range 3.5-5.5 mmol/L.

Glycated Hemoglobin Estimation
Glycated hemoglobin was estimated using kits by Teco diagnostics, Anaheim, USA. The kit employed a weak binding cation-exchange resin for the rapid separation of glycohemoglobin (fast fraction) from non-glycated hemoglobin (HbA o ). Normal range: <6%.

Serum cortisol Estimation
Serum and salivary cortisol was determined by the DRG cortisol enzyme immunoassay kit. It was obtained from DRG international Inc. USA. The DRG cortisol ELISA kit is a solid phase enzyme-linked immunosorbent assay (ELISA), based on the principle of competitive binding. Absorbance was read at 450nm with an ELISA microtiter plate reader. Normal range of serum cortisol -138 nmol/L-635 nmol/L.

Statistical Analysis
This was done using the PAWstatistic 18, a statistical package from SPSS Inc, California, USA. The results were expressed as Mean ± SD. The data was analyzed by Student's t-test and Analysis of variance (ANOVA) followed by a post hoc test using least significant difference (LSD). The level of significance was set at 95% confidence interval, where P-value less than .05 (P < .05) was considered as statistically significant. Correlation was done using Pearson's correlation. Graphs were done using Microsoft excel 2007 version.

Definition of Terms
Good glycaemic control: was defined as HbA1c value of < 7% in the diabetic patients.
Poor glycaemic control: was defined as HbA1c value of ≥7% in the diabetic patients [11].

Results
A comparison of the mean values of fasting plasma glucose, glycated haemoglobin and serum and salivary cortisol levels in diabetics and control subjects showed that the mean fasting plasma glucose, glycated haemoglobin and serum and salivary cortisol levels were significantly (P = .000) higher in diabetics when compared to the control subjects. There was no significant difference between the percentage difference between the serum and salivary cortisol levels in both the diabetics and controls Table 1. Table 2 shows a comparison of fasting plasma glucose, glycated haemoglobin, serum and salivary cortisol in diabetics and controls based on glycaemic control. The mean fasting plasma glucose, glycated haemoglobin and serum cortisol levels were significantly lower (P = .000) in controls when compared with the two groups of diabetics. A similar pattern was also observed in the salivary cortisol (P = .006). The diabetics with poor glycemic control had the highest levels of mean fasting plasma glucose (FPG) and glycated haemoglobin. However, serum and salivary cortisol levels between the diabetics with poor glycaemic control and those with good glycaemic control were similar. There was a significant positive correlation between serum cortisol and salivary cortisol in both the diabetic subjects (r = 0.362, P = .007) and controls (r =0.406, P = .019) (Figs. 1,2).  Mean ± SD Key: *-higher than that of diabetics with good glycaemic control; # -higher than that of controls

Discussion
In our study, the mean levels of both serum and salivary cortisol were significantly higher in type 2 diabetics when compared with controls. Cortisol is a glucocorticoid therefore increase in cortisol levels induces hyperglycemia. Its role in diabetes mellitus may however be adverse as it tends to sustain hyperglycemia leading to diabetesrelated complications. The values found in this study indicate that cortisol may be a contributor to the diabetic condition. This agrees with observations by Chiodini et al. [12] who reported that the degree of cortisol secretion as reflected by F24 was directly associated with both the type 2 diabetes and number of complications. The increased levels of cortisol in saliva of the diabetics is reflective of the increased levels of serum cortisol concentration in these patients.
Surprising however was the observation that there was no significant difference in the mean serum or salivary cortisol levels between poorly controlled diabetics and diabetics with good glycaemic control. This may imply that in people with diabetes there is altered cortisol secretion and/or metabolism irrespective of their glycaemic status. Findings by Oltmanns et al. [13] differ from those in this study. They reported that HbA1C was directly associated with cortisol secretion in type 2 diabetic subjects with normal HPA activity.
The salivary cortisol was about 70% lower than the serum cortisol in each group, this may be because salivary cortisol reflects or estimates free cortisol levels [14] rather total serum cortisol which was what was estimated in this study. Despite, this difference, the concentration patterns for serum and salivary cortisol in both diabetic groups and controls were similar. The percentage difference between serum and salivary cortisol in both groups did not vary significantly. There were also significant positive correlations between serum cortisol and salivary cortisol in both groups. This shows a consistency in the patterns of similarity observed in the groups with normal cortisol levels as well as those with high cortisol levels. This finding is similar to the observations made by Trilck et al [15] in the use of salivary cortisol in diagnosis of Cushing's syndrome and Marcus-Perlman et al [16] in the use of salivary cortisol in low dose ACTH test. The limitation of this research is that we did not estimate free serum cortisol. This might have given a clearer picture.

CONCLUSION
Cortisol levels in saliva reflected those in serum in both diabetics and controls therefore salivary cortisol may be used in place of serum cortisol in the management of diabetes.

CONSENT
All authors declare that written informed consent was obtained from the patient to participate in the study and publish the results.

ETHICAL APPROVAL
All authors hereby declare that all experiments have been examined and approved by the appropriate ethics committee and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.