Background

Thyroid disease and diabetes mellitus are common endocrine disorders encountered in clinical practice. Both diabetes mellitus and thyroid disorders mutually influence each other (American Diabetic Association 2014). Researches indicate that there is increased incidence of type 2 globally, highlighting increased incidence of childhood type 2 diabetes mellitus (Jin et al. 2011). Predictions for the future are that in 2030 the number of persons suffering type 2 and the percentage of individuals with the disease will be more than double (Wilds et al. 2004). In developing countries, increasing morbidity of type 2 is reported (Smithson 1998). The increasing incidence of type 2 is compounded by reports of the coexistence of thyroid dysfunction with diabetes (Udiong et al. 2015). Furthermore, disturbing is the report that the thyroid disorders may present with low or raised levels of thyroid hormones (Ghazali and Abbingesuku 2010; Gursoy and Turkel 1999; O’ Meara et al. 1993). Glycaemic index levels are commonly used for diagnosis and management of diabetes mellitus without due attention to the influence of thyroid hormone levels on those indices such as fasting plasma glucose (FPG), glycated haemoglobin (HbA1c) and C-peptide. Existing reports have shown that the associations between thyroid hormone and glycaemic indices are altered in diabetes mellitus (Ghazali and Abbingesuku 2010; O’ Meara et al. 1993), the general notion being that thyroid hormone levels fluctuate with FPG. The physiological and biochemical interrelationship between insulin and thyroid hormone and the influence of both insulin and iodothyronines on the metabolism of carbohydrates proteins and lipids have been extensively studied and recorded (Additional file 1).

Endogenous production of glucose is enhanced by insulin and thyroid hormones since both FT3 and FT4 stimulate GLUT4 messenger RNA and protein expression in skeletal muscle and elevate glucose uptake (Kemp et al. 1997). In hyperthyroidism, increased level of thyroid hormones leads to increase GLUT4 and consequently increase hepatic output of glucose resulting in hyperglycaemia. Increased lipolysis which also occurs in hyperthyroidism results in increased levels of free fatty acid (FFA) (Maxon et al. 1975). Moreover, in hyperthyroidism, the non-oxidative disposal of glucose is enhanced resulting in overproduction of lactic acid which enters the Cori cycle and promotes further gluconeogenesis (Ghazali and Abbingesuku 2010). Hyperthyroidism is also associated with raised levels of growth hormone, glucagon and catecholamine which further impair glucose metabolism (Niki et al. 1992; Tosi et al. 2012). On the other hand, hypothyroidism also affects glucose metabolism through several mechanisms and reduced glucose production is reported. There is reduced rate of glucose production in hypothyroidism, a consequence of decreased production of GLUT4 (Kemp et al. 1997). Hypothyroidism has been recognized as insulin-resistant state due to impaired insulin-stimulated glucose utilization in peripheral tissues.

This information has not defined the interrelationship between the glycaemic indices and the relationship between the individual thyroid hormones (T3, T4 and FT3) and the different glycaemic indices, namely FPG, HbA1c and C-peptide/insulin. This study evaluates the relationship between thyroid hormone levels and glycaemic indices in type 2 diabetic subjects with euthyroid, hyperthyroid, hypothyroid hormone levels and in non-diabetic control subjects.

Methods

Study location/subject selection

This was a cross-sectional study involving one hundred and seventy type 2 diabetic subjects attending University of Calabar teaching hospital, University of Calabar Medical Centre, Luciama Memorial Specialist hospital and Mediscience laboratories Ltd. to monitor their glucose levels as the diabetic subjects. They were all residents of Calabar, Cross River State, Nigeria. All diabetic subjects were confirmed type 2 who previously had or were having fasting plasma glucose concentrations above 6.1 mmol/l and were receiving treatment for the ailment. Apparently healthy age (38–60 years) and sex match non-diabetic volunteers also residing in Calabar served as control subjects all of whom had fasting plasma glucose levels below 6.1 mmol/l. All subjects were briefed on the objectives of the study verbally, and informed consent was obtained. Very ill diabetics and patients with diabetic complications on admission were excluded from the study. The subjects under study were divided into 5 groups as follows: all type 2, euthyroid, hypothyroid, hyperthyroid type 2 diabetics and non-diabetic subjects. The parameters selected for thyroid profile were thyroid-stimulating hormone: TSH, triiodothyronine T3, thyroxine, T4 and free thyroxine, FT4, which allowed assessment of TSH, FT4, T4 and T3 under the 5 groups studied.

Sample collection

Criteria for selection of type 2 diabetic subjects were: age of onset of diabetic as greater than 37 years, type of treatment the patient was receiving and their physician’s diagnosis on the patient records such diabetic subjects had fasting C-peptide levels of 0.38 ng/ml or above and were responding to treatment. Both the diabetic and non-diabetic control subjects were instructed to fast overnight (at least 8 h), before blood samples were collected from them. Ten millilitres of venous blood was collected from each subject and dispensed as follows: 2 ml into fluoride oxalate bottle for plasma glucose estimation, 2 ml into EDTA bottle for glycated haemoglobin (HbA1c) and the remainder into a plain bottle from where serum was harvested and stored frozen at − 20°C until use for thyroid hormones and C-peptide assays.

Laboratory methods

ELISA reagent kits obtained from Dslab Inc., California, USA, were used for the estimation of thyroid-stimulating hormone (TSH), total triiodothyronine (T3), total thyroxine (T4), free thyroxine (FT4) and C-peptide. Reagent kits obtained from UNICO Inc., USA, were used for the estimation of HbA1c, while Trinder glucose oxidase method (Trinder 1969) was used for the measurement of FPG.

Statistical analyses

Statistical analysis was performed using SPSS version 18.0. Pearson’s correlation was used for correlation analyses. Quality control of procedures was done using control sera at low, normal and raised levels as provided by the reagents’ manufacturers. The results obtained were expressed as mean ± SD. The level of significance was set at P < 0.05.

Results

FPG and HbAlc were positively correlated in all type 2 diabetic group (r = 0.211, P = 0.007) and in the hypothyroid type 2 diabetic subjects (r = 0.382, P = 0.011) and hyperthyroid groups (r = 0.295, P = 0.011). FPG also correlated positively with C-pep in all type diabetics (r = 0.224, P = 0.004) and hyperthyroid type 2 subjects (r = 0.481, P = < 0.001). There was no significant correlation between FPG and HbAlc or C-pep in the other group (Table 1). There were no significant correlations between FPG and all the thyroid hormones.

Table 1 Correlation analysis between FPG and C-pep, HbAlc and thyroid profile in all type 2 diabetics, euthyroid, hypothyroid, hyperthyroid diabetic and non-diabetic control subjects
Full size table

HbAlc correlated positively with TSH in all type 2 (r = 0.211, P = 0.012) and in hypothyroid subjects (r = 0.330, P = 0.031), and there was positive association between HbAlc and T4 in all types (r = 0.194, P = 0.038) and the hypothyroid groups (r = 0.379, P = 0.012). There were no observed correlations in other groups (Table 2).

Table 2 Correlation analysis between HbAlc and C-pep and thyroid hormones in all type 2, euthyroid, hypothyroid, hyperthyroid diabetics and non-diabetic subjects
Full size table

Table 3 shows a positive correlation between C-peptide and FT4 only in the hyperthyroid group (r = 0.279, P = 0.042) and negative correlation between C-pep and T4 in the euthyroid group. There were also significant positive correlations between C-peptide and T3 in all type 2 and euthyroid diabetic groups. C–peptide did not correlate with TSH in any of the five groups of subjects.

Table 3 Correlation analysis between C-peptide and thyroid hormones in all type 2, euthyroid, hypothyroid and hyperthyroid type 2 diabetics and non-diabetic subjects
Full size table

Discussion

The coexistence of both low and raised levels of thyroid hormones in type 2 diabetic subjects I has been reported in many studies (Gursoy and Turkel 1999; O’ Meara et al. 1993; Kemp et al. 1997; Maxon et al. 1975), but currently, glycaemic indices: FPG, HbA1c and C-peptide levels, are frequently used for the diagnosis, evaluation and management of diabetic subjects without due attention to the possible effects of the coexistence of low or raised thyroid hormone levels alongside type 2 diabetes in those subjects.

In this study, FPG correlated positively with HbAlc in all type 2, hypothyroid and hyperthyroid groups. Glycation is said to be a non-enzymatic catalysed reaction (Udiong et al. 2007), and thus, our current findings suggest that altered thyroid hormone levels may play a role in glycation in type 2 since there was no significant association in both non-diabetic subjects and euthyroid type 2 diabetic subjects. But the positive correlations between FPG and HbAlc were observed in both diabetic subjects with low and raised thyroid hormone levels prompting a conclusion that the positive association between FPG and HbAlc is not driven by abnormal thyroid hormone levels; otherwise, the direction of correlation would have been opposite. The correlations observed in all type 2, hypothyroid diabetics would then depend on the FPG levels and the percentage of subjects in those groups that had raised glucose levels and the period, the hyperglycaemia lasted (Dimitriadis et al. 1985).

FPG and C-peptide were positively correlated in all type 2 and hyperthyroid groups. There were no significant correlations between the two parameters in the euthyroid, hypothyroid diabetics and non-diabetic subjects. It may thus be reasoned that raised thyroid hormone levels enhance association between FPG and C-peptide and that the positive correlation found in all type 2 diabetic group is due to the presence of diabetes with raised thyroid hormone levels in that group. Raised thyroid hormone levels promote hyperglycaemia (Beer et al. 1989) and would aggravate existing hyperglycaemia in type 2. In type 2 diabetes subjects, the raised levels of FPG trigger increased secretion of C-peptide/insulin (Beer et al. 1989; Levin and Symth 1963). The correlation between FPG and C-peptide is enhanced by hyperthyroidism which increases glucose absorption in the gut (Levin and Symth 1963), alters pro-insulin processing (Matty and Seshadri 1963) and reduces the half-life of insulin by increasing the rate of degradation and enhancing release of biologically inactive insulin precursors (Arshad and Hussain 2013). FPG levels did not correlate with TSH, FT4, T4 or T3 levels in any of the 5 groups of subjects studied though positive correlation was reported between glucose and T4 in streptozotocin-induced diabetic rats (Hage et al. 2011), and the disparity may be due to differences in the cause and intensity of diabetes in diabetic humans and induced diabetic rats. Our results imply that thyroid hormones levels are not influenced by glucose levels or vice versa. HbAlc did not correlate with C-peptide in all type 2, hypothyroid, hyperthyroid and euthyroid diabetic and also in non-diabetic controls. Rather, glycated haemoglobin correlated positivity with TSH and T4 in all type 2 and hypothyroid diabetic groups. Raised TSH levels are associated with hypothyroidism and low TSH with hyperthyroid states (Lai et al. 2011) in non-thyroidal illness state such as type 2 mellitus; the findings are not different even in the presence of complicating factors.

Glycated haemoglobin also correlated positively with T4 in all type 2 and those with low thyroid hormone levels similar to the relationship between HbAlc and TSH even though TSH levels would normally be expected to be inversely related to T4 levels in primary thyroid disorders (Ghazali and Abbingesuku 2010). In type 2, where glucose, insulin and thyroid hormone levels are raised, there is also increased binding protein levels such thymoglobulin (TBG) level leading to raise total T4 levels. Prolonged hyperglycaemia prompted by diabetes or hyperthyroidism will give rise to raise HbAlc and correlation with T4 as observed in the study. Moreover, the rate of conversion of T4 to T3 is reduced or inhibited by raised levels of insulin found in type 2 resulting in raised T4 levels (American Diabetic Association 2014; Jin et al. 2011). There were no significant correlations between HbAlc and TSH or T4 in the euthyroid, hypothyroid and non-diabetic subjects. FT4 and T4 did not correlate with HbAlc in any of the 5 groups of subjects. This again shows that thyroid hormones do not influence glycation process, and hence, glycated haemoglobin results can be interpreted without reference to thyroid hormone levels.

C-peptide correlated positively with FT4 in hyperthyroid diabetic group, positively with T4 in the euthyroid diabetics and positively with T3 in all type 2 and euthyroid group. The associations observed between C-peptide and FT4, T4 and T3 are similar. T3 is biologically more potent than T4 (Niki et al. 1992), but in type 2, FT4 increases, while T3 decreases. In spite of the disparity in activity, the observed association between T4, T3 and C-peptide in diabetics with euthyroid hormones level shows that C-peptide levels are influenced by thyroid hormone levels. In type 2, insulin resistance leads to hyperglycaemia and increased C-peptide/insulin levels which may trigger thyroid hormone increase to enhance glucose metabolism, hence the correlation between C-peptide and the thyroid hormones (FT4 and T3). There was a positive correlation between C-peptide and T3 in all type 2 group, but the association between C-peptide and FT4 was not significant. The disparity may be due to the levels of and number of subjects in each of FT4 and T3 in the hyperthyroid group. Our study suggests that C-peptide levels are positively associated with thyroid hormones levels.

Conclusions

This study has shown that the presence of low or raised levels of thyroid hormones in the diabetic population influenced the levels of glycaemic indices in the general type 2 diabetic population. This implies that inclusion of thyroid hormonal profiling may be useful in proper diagnosis and management of diabetes mellitus.