THE IMPACT OF TSH LEVEL ON ALBUMINURIA AND GLOMERULAR FILTRATION RATE IN TYPE 2 DIABETES MELLITUS

Rasha M Abdel Samie M.D 1 ., Ula M. Al-Jarhi M.D 1 ., Mona M. Fathy 2 and Mona Mohsen M.D 2 . 1. Department of Internal Medicine, Cairo University. 2. Department of Chemical Pathology, Cairo University. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History


1234
Thyroid abnormalities can have a great impact on renal disease. [8] hypothyroidism is known to decrease sodium reabsorption by the kidney and to reduce renal blood flow. [9,10] Also, hypothyroidism is more prevalent in patients with impaired renal functions. [11] Patients with chronic kidney disease (CKD) who are euthyroid show better prognosis of renal disease as well as higher survival rates than those with inadequately controlled hypothyroidism. [12,13] The aim of this work was to investigate for the presence of an association between subclinical hypothyroidism and chronic kidney disease (CKD) in patients with type 2 diabetes mellitus.

Materials and Methods:-
Patients were recruited from Endocrinology and Diabetes outpatient clinic, Kasr Al Ainy Hospital Cairo University during the period November 2013 to February 2014. All patients were known to have type 2 diabetes mellitus, about 50% were on insulin treatment and the rest on oral hypoglycaemic drugs, all on regular follow up. Patients were consecutively enrolled into the study as long as they were 35-50 years old and with disease duration no more than 15 years, and had a normal range serum creatinine.
A written informed consent was obtained from all patients. Patients with known thyroid illness or clinically detected goiters were excluded from the study. Patients were screened for subclinical hypothyroidism by measuring serum TSH and free T4.

Laboratory Workup:-
The following labs were done to all patients: Lipid profile; triglycerides and total cholesterol by enzymatic colorimetric method and HDL by direct method in homogeneous phase, LDL was calculated by using the Friedewald formula), fasting plasma glucose level was measured by glucose oxidase method, serum creatinine was measured by kinetic colormetric assay, HbA1c was analyzed on Dimension (Siemens) based on a turbidimetric inhibition immunoassay (TINIA) principle, serum level of TSH and fT4 were analyzed on Cobas e411 (Roche) based on Electrochemiluminescence immunoassay, microalbumin / creatinine ratio (random urine sample) was analyzed using Bio systems kit based on latex turbidimetry.
Patients with overt hypothyroidism (with decreased free T4) were treated accordingly and excluded from the study. Patients with a decreased TSH beyond the normal range with any thyroid function pattern were excluded. The CKD-EPI creatinine equation was used to estimate eGFR. [14] TSH was treated as a continuous and categorical variable, which allowed for the comparison of the absolute amount of TSH to outcome variables as well as the comparison of outcome variables between SCH and non-SCH patients. Serum TSH was categorised as less than or "more than or equal to" 3.5 µIU/L to classify SCH, according to the variations in the TSH upper limit recommendation. [7] Statistics:-Analysis of data was done using SPSS (statistical program for social science version 12) as follows: Quantitative variables are expressed in mean +/-SD. Qualitative variables were expressed as number and percentage. Chi-square test was used to compare qualitative variables between groups. Unpaired t-test was used to compare quantitative variables, in parametric data (SD<50% mean). P value <0.05 was considered as significant. Univariate linear regression was used to test the effect of TSH as a continuous variable on A1c, A/C ratio and eGFR. A backward multivariable regression model was done to test the effect of various factors on albuminuria and eGFR.

Results:-
According to the selected criterion in this study, 88 patients were found to be euthyroid and 12 patients were found to have SCH. FT4 values among patients with SCH ranged from 0.72 -1.50 ng/dl, with mean ± SD, 1.12 ± 0.25 ng/dl. There was no difference in demographic data between the euthyroid group and those with SCH. The mean age was around 48 years for both groups (euthyroid group 48.4  6.8 years and in SCH 48.6  5.1 years, P=0.857 ). The duration of diabetes mellitus was also similar between the two groups, with the mean 8-9 years (euthyroid group 8.01  3.518 and in SCH 9.67  4559, P=0.312) . The study cohort included 80 female and 20 male patients. The percentage of female subjects in euthyroid group was 81.8% and in SCH 66.7%, P=0.194.
As regards glucose control, both groups were poorly controlled, especially the euthyroid group. FBS was more or less similar between both groups (euthyroid group 211  108 mg/dl and in SCH 221  112 mg/dl, P=0.657). However, mean HBA1c was higher in euthyroid group (9.5  1.75%) versus SCH group (8.7 ± 2.05%), P=0.118. There was no difference in lipid profile between the euthyroid group and those with SCH. The mean ± SD of serum triglycerides (TG) was 182.52 140.56 mg/dl in the euthyroid group and 181.17  89.44 mg/dl in SCH, P= 0.671. The mean ± SD of serum low density lipoproteins (LDL) was 142.81  44.40 mg/dl in the euthyroid group and 136.47  40.62 mg/dl in SCH, P= 0.681. The mean ± SD of serum high density lipoproteins (HDL) was 37.98  9.17 mg/dl in the euthyroid group and 40.42  9.95 mg/dl in SCH, P= 0.353.
Serum creatinine was significantly higher in the SCH group (0.98 ± 1.91 mg/dl) than euthroid group (0.65 ± 0.23 mg/dl), P=0.004. However, HBA1c, ACR and eGFR were not significantly different in both groups (Table 1). Univariate regression analysis, however, showed a significant association of TSH as a continuous independent variable with all three dependent variables: HBA1C, eGFR and ACR. Regression coefficients and P values are shown in Table 2.
A multivariable regression model was done to test the effect of gender, age, duration of diabetes, BMI, type of treatment, serum creatinine, FBS, HBA1c, TSH, TG, LDL and HDL on albuminuria and eGFR. No significant associations were found.

Discussion:-
SCH is an everlasting hot topic concerning determination of cut-off points for diagnosis and treatment. This study is an important extension to the ongoing research aiming at disease-specific individualisation of SCH diagnosis and treatment. This approach best addresses cost-effectiveness of SCH correction in relation to associated risks implicated by the condition. In this study, the increased risk of chronic renal disease (CKD) with SCH was sought in patients with type 2 diabetes mellitus.
The triad of increased serum creatinine, higher ACR and decreased eGFR was proven to be more frequent in patients with SCH as compared to euthyroid counterparts with type 2 diabetes in this study. In stages 2 and 3 of diabetic nephropathy, the GFR is elevated, and there is no rise in creatinine. Rising creatinine signifies a fall in GFR, which heralds the progression to stage 4 and ESRD. [15] This means that SCH may be an important factor in progression of diabetic nephropathy.
Although the measurement of albuminuria is the cornerstone for the diagnosis of diabetic nephropathy, there are some patients with either type 1 or type 2 diabetes who have decreased glomerular filtration rate (GFR) in the presence of normal urinary albumin excretion. For patients with type 2 diabetes in NHANES III (Third National Health and Nutrition Examination Survey; n = 1,197), low GFR (<60 ml · min−1 · 1.73 m−2) was present in 30% of patients in the absence of micro-or macroalbuminuria and retinopathy. Although renal biopsy was not performed, this observation was probably related to renal parenchymal disease other than classical diabetic glomerulosclerosis. These studies indicate that normoalbuminuria does not protect from a decrease in GFR in type 1 and type 2 diabetic patients. Therefore, GFR should be routinely estimated and albuminuria routinely measured for a proper screening of diabetic nephropathy. [15] This data suggest that the effect of SCH on diabetic nephropathy is not only exerted by increasing the severity of microalbuminuria or macroalbuminuria, but also by hastening the direct progression to stage 4. Review of previous studies in this field revealed a significant negative correlation between eGFR and TSH [16,17], a strong association between CKD and SCH [17][18][19][20][21], with increased prevalence of renal disease in SCH patients [17] in general.
However, besides a Japanese study done in 2014 with very similar results, no study has investigated the association between SCH and eGFR among patients with type 2 diabetes mellitus. In the latter study, SCH was not found to be an independent risk factor for CKD, but eGFR was lower in the SCH group than the euthyroid group. [22] This association between SCH and worsening renal function in patients with diabetic nephropathy highlights the importance of thyroxine replacement in such patients. A mouse model done in 2011 seemed promising. [23] 1236 Levothyroxine replacement was shown to preserve renal functions and independently predict renal outcome in 309 patients with stage 2-4 chronic kidney disease and SCH. [24] Another study showed that thyroid replacement therapy may slow down the deterioration of eGFR in a cohort of patients, 39% of which had diabetes mellitus. [25] Limitations:-The relatively small number of SCH subgroup (n=12) as compared with the euthyroid group (n=88), made it difficult to obtain significant results by multivariable analysis.

Conclusions and recommendations:-
Increased TSH was significantly associated with increased serum creatinine, increased ACR and decreased eGFR. SCH is thus regarded as a risk factor for diabetic nephropathy.
In view of the results of this study, the authors recommend that thyroid function screening should be added to routine follow-up of patients with type 2 diabetes mellitus. The cut-off for thyroid dysfunction in this subset of patients may fall to 3.5 µIU/L, if not lower. Levothyroxine replacement has to be carefully considered in such patients especially if eGFR was affected. Needless to point out, eGFR should be added to routine screening of diabetic nephropathy.

Futher research:-
Further studies should include prospective studies to test for the possible effect of thyroxine replacement on SCH patients as regards complications of diabetes.
Funding authors received no funding for this study Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.