Prevalence of hypothyroidism in patients with chronic kidney disease: a cross-sectional study from North India.

BACKGROUND: There is an increased prevalence of hypothyroidism in chronic kidney disease (CKD) patients as the glomerular filtration rate falls. However, there is a paucity of Indian data in this respect.
METHODS: A cross-sectional analysis was performed based on the database of the information system of a tertiary care hospital in northern India to retrieve results of nephrology CKD outpatients (> 18 years of age) from September 2013 to October 2015 to determine the prevalence of hypothyroidism in the non-dialysis-dependent CKD population. Overt hypothyroidism was defined by a thyroid-stimulating hormone (TSH) level > 5.5 mIU/L and free T4 level < 0.89 ng/dL with clinical symptoms. Subclinical hypothyroidism was defined by a TSH level > 5.5 mIU/L and a free T4 level ≥ 0.89 ng/dL.
RESULTS: Among 1,863 CKD patients, 358 patients underwent biochemical analysis for hypothyroidism. Among these, 143 had biochemical subclinical hypothyroidism and 59 had overt hypothyroidism. Patients in the overt hypothyroid group had significantly higher TSH levels and a lower free T4 level than those in the nonhypothyroid group. Patients with hypothyroidism (both clinical and subclinical) had significantly lower serum albumin and serum calcium levels than those in the nonhypothyroid group. Intact parathyroid hormone was also significantly higher in the hypothyroid groups. An increased prevalence of hypothyroidism was observed in patients with a reduction in the glomerular filtration rate.
CONCLUSION: There is growing evidence of increased prevalence of hypothyroidism in dialysis-independent CKD patients. A number of findings such as lower serum albumin, serum calcium, and hemoglobin levels and higher intact parathyroid hormone levels are seen in this group. Specific treatment can help improve these. Hence, there is a need to formulate guidelines to screen this population for hypothyroidism.

Introduction

It has been shown that in chronic kidney disease (CKD), as the glomerular filtration rate (GFR) falls, there is a higher possibility of developing clinical and subclinical hypothyroidism (SCH) [1]. With falling GFR, there are a number of abnormalities developing in the thyroid gland at both structural and functional level. There is an increase in thyroid volume as the GFR falls [2]. Low T3 syndrome is also commonly seen in CKD population, which is possibly an adaptation of the chronic inflammatory and malnourished state prevalent in these patients. The overlapping features of CKD and hypothyroidism make it all the more challenging for the clinician to timely diagnose and treat it. In spite of the growing volume of information, there is a dearth of Indian data with respect to SCH and overt hypothyroid prevalence in patients with CKD who are not dialysis dependent. To gain further insight into this, we conducted a survey of the CKD population visiting this tertiary care center to determine the prevalence of hypothyroidism in the non–dialysis-dependent CKD population.

Methods

A cross-sectional analysis based on the database of the information system of a tertiary care hospital in northern India to retrieve results of nephrology CKD outpatients (> 18 years of age) referred to this tertiary care center from September 2013 to October 2015 was conducted.

Serum thyroid-stimulating hormone (TSH) and free T4 (FT4) concentrations were quantified by direct chemiluminescence using acridinium ester technology on the ADVIA Centaur XP analyzer (Siemens Healthcare, Tarrytown, New York, USA). Functional sensitivity for TSH and FT4 was quoted by the manufacturer as 0.004–150 mIU/L and 0.1–12.0 ng/dL, respectively. Reference values in our laboratory were 0.35–5.5 mIU/L for TSH and 0.89–1.78 ng/dL for FT4, respectively. The estimated GFR (eGFR) was calculated using the 2009 chronic kidney disease Epidemiology Collaboration creatinine equation, and the stages of CKD were defined according to the kidney disease improving global outcome guidelines for evaluation of CKD.

Overt hypothyroidism would be defined by a TSH level > 5.5 mIU/L and a FT4 level < 0.89 ng/dL with clinical symptoms. SCH would be defined by a TSH level > 5.5 mIU/L and a FT4 level ≥ 0.89 ng/dL (the lower limit of the normal range).

Inclusion criterion was all CKD patients (> 18 years of age) not requiring long-term dialysis with TSH levels > 5.5 mIU/L.

Exclusion criteria were as follows: subjects younger than 18 years, pregnant women, subjects receiving concurrent treatment with drugs that could contribute to hypothyroidism, and subjects receiving antithyroid drugs presumably for hyperthyroidism.

All secondary cases of hypothyroidism and subjects in whom kidney functions could not be estimated because of missing serum creatinine values or those in whom TSH or FT4 levels were not available were also excluded.

Statistical analysis

Data were analyzed using SPSS 16 (IBM Corporation, New York, USA). The Student t test, analysis of variance (one way), Mann–Whitney test for scattered data were performed, and odds ratio with 95% confidence interval is reported. A P value < 0.05 was considered significant.

Results

Study population

A total of 1,863 adult participants had valid serum creatinine measurements and eGFR. Among 1,863 CKD patients seen in the nephrology out patient department, 358 patients underwent biochemical analysis for hypothyroidism. Among these, 156 had normal TSH values (between 0.35 and 5.5 mIU/L), whereas 143 had biochemical SCH (i.e., TSH > 5.5 mIU/L with normal FT4 levels), and 59 had overt hypothyroidism (i.e., TSH > 5.5 mIU/L with FT4 levels < 0.89 ng/dL). The prevalence of subclinical and overt primary hypothyroidism together in the total CKD population was 10.84%, whereas it was 56.42% in the subjects tested for hypothyroidism (Fig. 1 and Table 1).

Figure 1

Flow chart of study design.

CKD, chronic kidney disease; FT4, free T4; OPD, out patient department; TSH, thyroid-stimulating hormone.

Table 1

Prevalence of subclinical and overt hypothyroidism in different stages of GFR

GFR	Primary hypothyroidism (n = 59), N (%)	Subclinical hypothyroidism (n = 143), N (%)	Nonhypothyroid (n = 202), N (%)
2	4 (6.78)	16 (11.19)	20 (12.82)
3a	1 (1.69)	14 (9.79)	14 (8.97)
3b	6 (10.18)	25 (17.48)	48 (30.77)
4	24 (40.68)	41 (28.67)	26 (16.68)
5	24 (40.68)	47 (32.84)	48 (30.76)
Total	59	143	202

GFR, glomerular filtration rate.

The mean age was 55.89 ± 12.85 and 55.16 ± 14.28 years in the subclinical and overt hypothyroidism groups, respectively.

Characteristics of persons with and without prevalent hypothyroidism

Overall, individuals with clinical hypothyroidism and SCH were roughly of the same age (Table 2). In gender distribution between the groups, the number of men was higher in the overt hypothyroid group although nonsignificant. In the subclinical hypothyroid group, both the sexes were almost equally distributed. Systolic blood pressure was significantly higher in overt and subclinical hypothyroid groups compared to the nonhypothyroid group (P < 0.001).

Table 2

Comparison of different variables in overt hypothyroidism, subclinical hypothyroidism, and nonhypothyroids

Variable	Primary hypothyroidism (n = 59)	Subclinical hypothyroidism (n = 143)	Nonhypothyroid (n = 156)
Age (y)	55.2 ± 14.3	55.9 ± 12.9	55.0 ± 14.4
Sex (M, F)	37, 22	71, 72	112, 44
Comorbidities
Diabetes mellitus	18 (30.51)	83 (58.04)	70 (44.87)
Hypertension	43 (72.88)	127 (88.81)	120 (76.92)
CAD	3 (5.08)	4 (2.79)	4 (2.56)
COPD	0	3 (2.1)	2 (1.28)
CVA	0		2 (1.28)
BMI	25.2 ± 4.4	24.1 ± 7.8	23.0 ± 5.8
SBP (mmHg)	147.2 ± 24.6	157.0 ± 26.3	134.0 ± 16.5
DBP (mmHg)	84.5 ± 11.6	88.0 ± 7.7	88.0 ± 9.4
TSH	83.0 ± 105.0	9.0 ± 4.4	2.0 ± 0.9
Free T4	0.7 ± 0.1	1.2 ± 0.1	1.3 ± 0.2
Triglyceride (mg/dL)	168.5 ± 103.8	153.0 ± 86.4	169.2 ± 95.6
Total cholesterol (mg/dL)	141.9 ± 74.5	159.7 ± 59.9	161.4 ± 48.4
Hemoglobin (g/dL)	9.6 ± 2.3	9.7 ± 1.9	10.1 ± 2.2
Serum calcium (mg/dL)	8.3 ± 0.9	8.7 ± 0.9	8.9 ± 0.8
Serum albumin (g/dL)	3.7 ± 0.7	3.6 ± 0.6	3.8 ± 0.5
Phosphorus (mg/dL)	5.2 ± 2.2	4.6 ± 1.6	5.0 ± 1.3
SALP (U/L)	141.2 ± 64.7	145.6 ± 82.6	168.3 ± 51.1
iPTH (ng/L)	236.1 ± 195.6 (median = 156)	183.5 ± 190.4 (median = 161.6)	39.2 ± 80.2 (median = 23.67)
Urine proteinuria (mg/dl)	134.8 ± 129.8 (median = 100)	113.2 ± 131.9 (median = 100)	79.5 ± 96.5 (median = 30)

Data are presented as n (%) or mean ± SD.

BMI, body mass index; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; CVA, cerebrovascular accident; DBP, diastolic blood pressure; iPTH, intact parathyroid hormone; SALP, serum alkaline phosphatase; SBP, systolic blood pressure; SD, standard deviation; TSH, thyroid-stimulating hormone.

FT4 concentrations differed significantly among the overt, subclinical hypothyroid, and nonhypothyroid groups (0.7 ± 0.1, 1.2 ± 0.2, and 1.3 ± 0.2 mIU/L, respectively, P = 0.001). Mean spot urine protein of 134.8 ± 129.8 mg/dL (median 100 mg/dL) was significantly higher in the overt hypothyroid group (P < 0.02). Patients with hypothyroidism (both clinical and subclinical) had significantly lower serum albumin (P = 0.00) and serum calcium levels (P = 0.002) than those in the nonhypothyroid group. Intact parathyroid hormone (PTH) was also significantly higher in the hypothyroid groups than in the nonhypothyroid group (P = 0.001). Hemoglobin levels were lower in the hypothyroid groups than in the nonhypothyroid group although it was not significant (P = 0.068). The patients in the hypothyroid groups (both clinical and subclinical) also had higher BMI compared with patients in the nonhypothyroid group (24.7 ± 6.5 vs. 23.0 ± 5.8). No significant difference was seen in the lipid profile among all the groups.

Among comorbidities, hypertension was the most common followed by diabetes mellitus across all groups.

Kidney function and prevalent hypothyroidism

We observed that the prevalence of hypothyroidism was increased in subjects with reduced GFR, ranging from 32.68% for persons with eGFR ≥ 30 mL/min/1.73 m2 to 67.32% in patients with eGFR < 30 mL/min/1.73 m2. Compared with GFR ≥ 60 mL/min/1.73 m2, the odds of hypothyroidism increased among subjects with lower eGFR: adjusted odds ratio, 1.07 (95% confidence interval, 0.41–2.78) for an GFR of 45–59 mL/min/1.73 m2, 0.65 (0.29–1.39) for an GFR of 30–44 mL/min/1.73 m2, 2.5 (1.16–5.39) for an GFR of 30–15 mL/min/1.73 m2, and 1.48 (0.72–3.74) for an GFR of < 15 mL/min/1.73 m2.

Discussion

Thyroid autoimmunity and subclinical primary hypothyroidism are highly prevalent in CKD patients not requiring long-term dialysis treatment [3]. In their study, Lo et al [2] reported a prevalence of hypothyroidism of 23.1% in CKD patients with an eGFR < 30 mL/min/1.73 m2. In another study, subclinical hypothyroidism and clinically apparent hypothyroidism have been reported to occur in ∼18–20% of patients with CKD not requiring renal replacement therapy [1]. This study differs from these previous observations by demonstrating a prevalence of subclinical and overt primary hypothyroidism (10.8%). An increased prevalence of subclinical and overt primary hypothyroidism in persons with reduced eGFR independent of age and gender was seen in this study. This is in line with the observation made by Chonchol et al [1]. In this study, the absolute prevalence of hypothyroidism in the lower GFRs is higher than that reported in other studies, which may be due to the smaller sample size in the present study. Majority of the patients in this study fell in the CKD stage 4/5 category, which could be due to the fact that most of the CKD patients referred to this tertiary care center have a low GFR. A multicentric screening program for hypothyroidism in the CKD population can provide a better picture of its actual prevalence.

Higher TSH levels are seen with increasing age [4]. The mean age in this study was 55.2 ± 14.3 and 55.9 ± 12.9 years in subclinical and overt hypothyroidism groups, respectively.

The TSH level is often elevated in CKD in response to thyrotropin from pituitary as a result of uremic effect [5]. TSH also loses its circadian rhythm along with compromised bioactivity due to poor glycosylation. Chronic metabolic acidosis has also been labeled as one of the contributing factors in the rise of hypothyroidism cases in CKD population [6]. The Wolff–Chaikoff effect [7] has been cited as a causative phenomenon behind the rise of this disorder in diabetic kidney disease patients. In our study, diabetes mellitus was present in 18 (30.51%) and 83 (58.04%) patients with overt hypothyroidism and SCH, respectively, although it was not significantly different from the nonhypothyroid group.

Dyslipidemia is seen throughout the spectrum of thyroid dysfunction although it is of much milder degree with TSH levels between 5 and 10 mU/L compared to TSH > 10 mIU/L [8], [9], [10]. A few reports [11], [12] have shown significantly elevated total cholesterol with TSH < 10 mIU/L in comparison to euthyroids. Contrary to this, total cholesterol and triglyceride levels were not significantly different in hypothyroids (overt and subclinical) compared to nonhypothyroids in our study.

In agreement to our study, Shantha et al [13] have also reported lower serum albumin levels in this population. Higher degree of proteinuria in the hypothyroid group can explain this finding.

An increase in TSH levels may lead to increase in PTH levels either due to thyrotropin-releasing hormone stimulation or error in measurement by immunometric assays [14]. Our study followed the trend. Hypothyroidism leads to a reduction of osteoclast bone reabsorption and osteoblast formation, slowing the remodeling process and increasing the time taken in the remodeling cycle, mainly due to the prolongation of the mineralization phase. A slight increase in bone mass may occur, albeit not with a reduced risk of fracture.

The incidence of anemia in hypothyroidism ranges between 23% and 60%. It is generally multifactorial. Hypothyroidism induces a decrease in bone marrow volume along with a decrease in nucleated blood cell precursors [15], [16], [17]. A decrease in gastric acid production in the hypothyroid state leads to a fall in iron absorption causing iron-deficiency anemia [18]. Increased blood loss due to menorrhagia may also lead to iron deficiency. Folate deficiency and higher incidence of pernicious anemia [19] may lead to macrocytic anemia. In our study, hemoglobin levels were also lower in subclinical and hypothyroid groups but not independent of GFR.

Our study also had several limitations. This is a cross-sectional study, so causality cannot be established. As this study was not based on screening, our observations may have underestimated the true prevalence of hypothyroidism in CKD. In addition, our study lacks details about the possible etiology of SCH or a measure of antithyroid antibodies. A prospective study is desirable, which can throw light on the effect of thyroxine treatment on GFR and give a better picture of the true prevalence of hypothyroidism in CKD population.

Conflicts of interest

The author has no conflicts of interest to declare.