Low thyroid function is not associated with an accelerated deterioration in renal function

Low thyroid function is not associated with an accelerated deterioration in renal function Abstract Background Chronic kidney disease (CKD) is frequently accompanied by thyroid hormone dysfunction. It is currently unclear whether these alterations are the cause or consequence of CKD. This study aimed at studying the effect of thyroid hormone alterations on renal function in cross-sectional and longitudinal analyses in individuals from all adult age groups. Methods Individual participant data (IPD) from 16 independent cohorts having measured thyroid stimulating hormone, free thyroxine levels and creatinine levels were included. Thyroid hormone status was defined using clinical cut-off values. Estimated glomerular filtration rates (eGFR) were calculated by means of the four-variable Modification of Diet in Renal Disease (MDRD) formula. For this IPD meta-analysis, eGFR at baseline and eGFR change during follow-up were computed by fitting linear regression models and linear mixed models in each cohort separately. Effect estimates were pooled using random effects models. Results A total of 72 856 individuals from 16 different cohorts were included. At baseline, individuals with overt hypothyroidism (n = 704) and subclinical hypothyroidism (n = 3356) had a average (95% confidence interval) −4.07 (−6.37 to −1.78) and −2.40 (−3.78 to −1.02) mL/min/1.73 m2 lower eGFR as compared with euthyroid subjects (n = 66 542). In (subclinical) hyperthyroid subjects (n = 2254), average eGFR was 3.01 (1.50–4.52) mL/min/1.73 m2 higher. During 329 713 patient years of follow-up, eGFR did not decline more rapidly in individuals with low thyroid function compared with individuals with normal thyroid function. Conclusions Low thyroid function is not associated with a deterioration of renal function. The cross-sectional association may be explained by renal dysfunction causing thyroid hormone alterations. chronic renal failure, CKD, creatinine clearance, epidemiology, thyroid function INTRODUCTION The prevalence of chronic kidney disease (CKD) is globally increasing, reaching endemic levels [1]. Since this growth is accompanied by a substantial increase in cardiovascular morbidity and mortality [2, 3], prevention of CKD and its secondary complications are of increasing importance. To date, however, aggressive management of known risk factors such as blood pressure, albuminuria and glucose control has not resulted in a clear reduction of this trend. These observations stress the need for further studies on other risk factors being amenable for treatment. In cross-sectional studies, lower renal function is accompanied by reductions in free thyroxine (fT4) and triiodothyronine (T3) and an elevation in serum thyroid stimulating hormone (TSH) levels [4–6]. This finding can be interpreted in two ways: first, CKD (similar to other chronic illnesses) induces a systemic lowering of the hypothalamic–pituitary–thyroid (HPT) axis, known as ‘non-thyroidal illness’ [7]. Alternatively, primary hypothyroidism could be the cause of a reduction in renal function. Indeed, studies in patients with severe primary hypothyroidism show a consistent reduction in renal function, which resolves after initiation of thyroid hormone supplementation [8, 9]. Although large-scale observational studies show clear associations between subclinical hypothyroidism and an increased risk for heart failure [10], coronary heart disease and mortality [11] findings are not consistent with the association between subclinical hypothyroidism and lower renal function [12–14]. In an observational study in patients with CKD Stages 2–4 and subclinical hypothyroidism, subjects not being prescribed thyroid hormone treatment showed a more rapid decline of renal function as compared with patients who received thyroid hormone supplementation [12, 13]. However, no association between thyroid hormone status and a decline in renal function was observed in a population-based study of the oldest old [14]. In light of these conflicting findings, this study sets out to evaluate the association between thyroid hormone status and renal function cross-sectionally and longitudinally by performing an individual patient data (IPD) meta-analyses on data from 16 independent cohorts participating in the Thyroid Studies Collaboration. MATERIALS AND METHODS Data from 16 different cohorts (four from the Netherlands, three from Italy, two from USA, two from Japan (and partly Brazil), and one from Germany, Norway, Belgium, Australia and Ireland), providing measures of thyroid and renal function were used for our analyses [11]. Nine of these cohorts [15–23] were also included in an earlier study evaluating the association between thyroid function and cardiovascular mortality [11]. Details of these cohorts have been described previously. In addition to these nine cohorts, seven cohorts [24–30] with data on thyroid and renal function were added to the collaboration. Six out of these seven cohorts were population-based studies; two comprising individuals from all age categories [27, 30], two having included those with an average age ∼70 years [28, 29] and two other studies included specifically the oldest [25, 26]. The seventh study was comprised of patients with chronic heart failure [24]. Thyroid function Thyroid function tests were measured at baseline in each cohort. We used common definitions to define thyroid hormone groups by using cohort-specific cut-off values which are summarized in Supplementary data, Appendix S1: (i) overt hypothyroidism was defined as elevated TSH levels in combination with reduced fT4 levels; (ii) subclinical hypothyroidism was defined as an elevated serum TSH level with a normal fT4 concentration; (iii) subjects were categorized in the euthyroid group when having TSH levels within the specific reference range; and (iv) those with lowered TSH levels with or without elevated fT4 levels were categorized as (subclinically) hyperthyroid. Subjects with subclinical hyperthyroidism and overt hyperthyroidism were combined because of low numbers in each group. Renal function Creatinine levels were measured according to each cohort’s protocol. Differences exist in their methodology; five cohorts utilized colorimetric assessments [15–17, 22, 25], five cohorts utilized the traditional Jaffé method [19, 21, 26, 27, 29] and the newer enzymatic method was applied in four cohorts [20, 23, 28, 30]. In the Bari cohort [24], creatinine levels were measured with both the Jaffé and colorimetric method. In the Nord-Trøndelag Health (HUNT) study [18], the baseline examination (HUNT2) was performed using the Jaffé method and subsequently adjusted by means of a validated calibration formula [31]. The alkaline picrate methodology was used for the follow-up examination (HUNT3) . Estimated glomerular filtrations rate (eGFR) was assessed by means of the four-variable Modification of Diet in Renal Disease (MDRD) formula [32]. For sensitivity analyses, eGFR was calculated on basis of the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula [33]. The MDRD formula was used as primary outcome instead of the CKD-EPI because not all creatinine measurements were based on traceable isotope dilution mass spectrometry [34]. Statistical analyses Baseline characteristics for each cohort are presented as means with SDs or numbers with percentages (%), as appropriate. To analyse the association between thyroid hormone status and renal function, a two-stage IPD meta-analysis was used as previously specified [11]. First, effect estimates were calculated at a cohort level. Thereafter, they were pooled at a meta-analysis level. For the cross-sectional associations between thyroid hormone status and renal function, linear regression analyses were fitted. Thyroid hormone groups were entered as categorical variables with the euthyroid group serving as the reference. Effect estimates (betas) represented the difference in eGFR (mL/min/1.73 m2) at baseline for the specific thyroid hormone group with respect to the euthyroid group. The same concept was applied to TSH and fT4 groups, again those with normal levels serving as the reference category. TSH and fT4 were also entered as continuous variables in which effect estimates illustrated an increase in eGFR per 1 mIU/L and per 1 pmol/L increase in serum TSH and fT4 levels, respectively. For the longitudinal analyses, examining the association between thyroid hormone status at baseline and the change in renal function over time, linear mixed models were fitted in each cohort separately. Because of the large variability in the number and timing of measurement points within cohorts and between cohorts, we chose to adopt random effects models. An average change in eGFR (mL/min/1.73 m2/year) for each cohort was calculated and presented in a figure. Slopes were not pooled because of large heterogeneity in effect estimates and participant characteristics. As for the linear regression analyses, thyroid hormone groups were entered as categorical variables. In addition, an interaction term of thyroid hormone group and time was included to allow for dependence of the slope on thyroid hormone status. Effect estimates obtained from these models indicated the additional change in eGFR per year as compared with the change in the euthyroid group. All models were adjusted for age, sex, cardiovascular disease and when available, for thyroid hormone supplementation and/or anti-thyroid medication. In sensitivity analyses, models were rerun also adjusting for diabetes mellitus, if available. Outcomes obtained from linear regression analyses (cross-sectionally) and from linear mixed models (longitudinally) were pooled by means of random effects models assuming the variance model as proposed by DerSimonian and Laird [35]. Sensitivity analyses were performed by rerunning all previous models in subgroups of sex and age. Another sensitivity analysis was performed excluding the HUNT study because creatinine measurements at baseline and follow-up were performed with different assays. The same was done for the Health, Aging and Body Composition (Health ABC) study. Also, sensitivity analyses were done by excluding all cohorts with positive changes in eGFR over time [22, 26, 27, 29]. To further examine the potential of selection/publication bias, funnel plots were created. Bubble plots were created plotting the effect size against mortality rates in each cohort. For differences, a 95% confidence interval (CI) not including zero was considered to indicate statistical significance. For all other tests, a P-value <0.05 was adopted as cut-off. Stata 12.1 (StataCorp LP, Texas, USA ) was used to perform all analyses. Figures were created using Stata and Prism 5.02 (GraphPad Software Inc., La Jolla, CA, USA; 1992). RESULTS This study included data from 16 cohorts, comprising a total of 72 856 individuals of whom 704 were hypothyroid, 3356 subclinically hypothyroid, 66 542 euthyroid and 2254 (subclinically) hyperthyroid. Baseline characteristics of the different cohorts are presented in Table 1. As illustrated, the average age at baseline ranged from 49 to 85 years and the proportion of individuals with pre-existing cardiovascular disease from 1.9% to 100%. Within the different cohorts, 0–9.9% of subjects were prescribed thyroid hormone replacement therapy and 0–4.7% used antithyroid medication. Usage of thyroid hormone supplementation was more common in the subclinical hypothyroid and hypothyroid groups, whereas more individuals in the hyperthyroid group used antithyroid medication (Supplementary data, Appendix S2). Table 1. Baseline characteristics of the different cohorts Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  a Combined thyroid supplementation/antithyroid medication. CHS, Cardiovascular Health Study; EPIC, European Prospective Investigation into Cancer and Nutrition Study; SHIP, Study of Health in Pomerania; InCHIANTI, Invecchiare in Chianti; PREVEND, Prevention of Renal and Vascular End-Stage Disease; na, not available; subcl, subclinical; hypo, hypothyroidism. Table 1. Baseline characteristics of the different cohorts Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  a Combined thyroid supplementation/antithyroid medication. CHS, Cardiovascular Health Study; EPIC, European Prospective Investigation into Cancer and Nutrition Study; SHIP, Study of Health in Pomerania; InCHIANTI, Invecchiare in Chianti; PREVEND, Prevention of Renal and Vascular End-Stage Disease; na, not available; subcl, subclinical; hypo, hypothyroidism. Cross-sectional analyses In Figure 1, the average eGFR per cohort is shown with mean (SD) values ranging from 59.0 (14.4) mL/min/1.73 m2 in the Leiden 85-plus study to 102.6 (26.7) mL/min/1.73 m2 in Radiation Effects Research Foundation (RERF). Figure 2 details forest plots presenting the differences in eGFR at baseline between the hypothyroid, subclinical hypothyroid and subclinical hyperthyroid group versus euthyroid group for each cohort separately. Pooled estimates show that eGFR was on average −4.07 (95% CI: −6.37 to −1.78) mL/min/1.73 m2 lower in the hypothyroid and −2.40 (−3.78 to −1.02) mL/min/1.73 m2 lower in the subclinical hypothyroid group as compared with the euthyroid group. Conversely, average eGFR was 3.01 (1.50–4.52) mL/min/1.73 m2 higher in (subclinical) hyperthyroid subjects. FIGURE 1: View largeDownload slide Mean (SD) eGFR (mL/min/1.73 m2) at baseline in the different cohorts. FIGURE 1: View largeDownload slide Mean (SD) eGFR (mL/min/1.73 m2) at baseline in the different cohorts. FIGURE 2: View largeDownload slide Forest plots providing the pooled differences in eGFR (mL/min/1.73 m2) at baseline in each thyroid group as compared with the euthyroid group. FIGURE 2: View largeDownload slide Forest plots providing the pooled differences in eGFR (mL/min/1.73 m2) at baseline in each thyroid group as compared with the euthyroid group. FIGURE 2: View largeDownload slide Continued. FIGURE 2: View largeDownload slide Continued. Longitudinal analyses Of the 16 cohorts, 13 contributed a total of 113,670 measurements during 329 713 patient-years of follow-up. Figure 3 depicts adjusted average annual changes in eGFR (mL/min/1.73 m2/year) per cohort, showing a range in annual change from −1.43 (0.06) to + 8.98 (1.05) mL/min/1.73 m2/year. Figure 4 demonstrates the pooled differences in eGFR change per year within the different thyroid hormone groups as compared with the euthyroid group. The change in eGFR was 0.35 (0.07–0.64) mL/min/1.73 m2 per year higher in the overt hypothyroid compared with the euthyroid group. No significant differences in eGFR change over time were noted in the other thyroid hormone groups as compared with euthyroid subjects. FIGURE 3: View largeDownload slide Mean (SD) changes in eGFR (mL/min/1.73 m2/year) in the different cohorts. FIGURE 3: View largeDownload slide Mean (SD) changes in eGFR (mL/min/1.73 m2/year) in the different cohorts. FIGURE 4: View largeDownload slide Forest plots providing the pooled additional changes in eGFR (mL/min/1.73 m2/year) per year in each thyroid group as compared with the euthyroid group. n = number of observations. FIGURE 4: View largeDownload slide Forest plots providing the pooled additional changes in eGFR (mL/min/1.73 m2/year) per year in each thyroid group as compared with the euthyroid group. n = number of observations. FIGURE 4: View largeDownload slide Continued. FIGURE 4: View largeDownload slide Continued. All analyses were repeated cross-sectionally and longitudinally in subgroups of sex (Supplementary data, Appendices S3 and S5) and age (<50, 50–65, 65–80 and >80 years) (Supplementary data, Appendices S4 and S6) rendering no differences in results. When models were adjusted for diabetes mellitus, results were not substantially different (Supplementary data, Appendices S7 and S8). Using the CKD-EPI formula instead of the four-variable MDRD formula, findings were comparable (data not shown). When pooling was repeated excluding the HUNT and Health ABC studies, similar results were found. The same held for the analyses excluding cohorts with positive changes in eGFR over time. Finally, analyses were repeated in individuals not using thyroid hormone supplementation or antithyroid medication, with no effect on the findings (data not shown). To examine the possibility of selection bias as a consequence of death, study-specific effect estimates were regressed on mortality incidence rates for each study, showing no statistically significant association (data not shown). Finally, a funnel plot did not show evidence of publication bias (data not shown). DISCUSSION In this IPD meta-analysis comprising data from 72 856 individuals out of 16 independent cohorts, we found a positive cross-sectional association between thyroid function and renal function in which those with low thyroid function had lower eGFR values as compared with euthyroid and (subclinical) hyperthyroid subjects. During follow-up, low thyroid function was not associated with an additional decline in renal function as compared with the euthyroid group. The presence of a cross-sectional association between low thyroid function and renal dysfunction aligns with findings in several previous cohorts [14, 36, 37], two of which were also included in this meta-analysis [14, 36]. Also compatible with our previous report, low thyroid function did not associate with an additional decline in renal function versus a euthyroid state [14]. Rather, we observed a relative increase in renal function in subjects with overt hypothyroidism as compared with individuals with thyroid hormone concentrations within the reference range. Three potential clarifications should be considered: first, whereas those with hypothyroidism have a lower eGFR at baseline, the observation of an increase in eGFR values over time relative to the euthyroid group may be explained by the concept of ‘regression to the mean’. Regression to the mean implies that when a variable has extreme values at a certain measurement, a second measurement will tend, by chance, to show a value closer to the true mean. In case of an extremely high value, the second measurement tends to be lower and vice versa, in case of an extremely low value, the second measurement will tend to portrait a higher value. Secondly, as overt hypothyroidism is generally considered an indication for thyroid hormone supplementation, the relative rise in eGFR over time could be the resultant of treatment rather than due to hypothyroidism itself. Yet, sensitivity analyses excluding subjects on thyroid medication did not reveal differential findings. Nevertheless, changes in thyroid status and medication during follow-up could have translated into different outcomes. Finally, patients with hypothyroidism could be more inactive resulting in a lower muscle mass, thereby lowering creatinine levels. We did not observe a difference in eGFR change over time in the subclinical hypothyroid group as compared with euthyroid subjects. Because subclinical hypothyroidism is not a strict indication for thyroid hormone supplementation, thyroid hormone supplementation likely did not play an interfering role. Results of a previous non-randomized study suggest that thyroxine supplementation preserves renal function over time in patients with CKD Stages 2–4 [12]. Findings from that study are, however, hampered by several limitations. First, in the absence of a randomized design and/or appropriate adjustment, confounding by indication may have imposed systematic error. Patient characteristics and physician preferences likely influenced the decision to initiate treatment. The non-treatment group indeed seemed overall less healthy than the treatment group. Secondly, 49 out of 358 individuals were excluded from the analyses because of a follow-up duration <12 months. This loss to follow-up may have been dependent on treatment status and outcome, and as a result, have introduced selection bias. Therefore, in addition to previous literature, current findings do not support a causal relationship between subclinical hypothyroidism and a decline in renal function over time. It is of interest to speculate why eGFR increased over time in cohorts with an average higher age [25, 26]. One of the explanations may be that those individuals with more rapid declines in renal function died sooner. Alternatively, as muscle metabolism and habitus change in old age, conventional equations for assessing eGFR are poorly validated in elderly individuals [38]. Nevertheless, in the Leiden 85-plus study, comprising individuals in similar age categories as PROSPER (Prospective Study of Pravastatin in the Elderly at Risk) and BELFRAIL (the Belgian cohort of the Very Elderly), average annual change in eGFR was considerably lower. When repeating our analyses excluding cohorts with positive changes in eGFR, results did not change, leading us to believe that this paradoxical increase in eGFR would not have translated into bias. Given the absence of a longitudinal association, the concept of reversed causation (CKD causing thyroid hormone abnormalities) may explain the observed cross-sectional association between low thyroid and renal function in our study. CKD, and especially end-stage renal disease, is frequently accompanied by abnormal TSH, low triiodothyronine (fT3) and fT4 levels fitting the spectrum of so-called ‘non-thyroidal illness’ [39]. In the absence of primary disease in the HPT axis, its pathogenesis is multifactorial and occurs at multiple levels including peripheral deiodinase-dependent conversion defects and central alterations in thyroid hormone signalling [40]. It could be speculated that deiodinase defects in early phases prevail over central mechanisms, leading to a compensatory increase in TSH secretion. Further studies on this hypothesis could include (free) T3 measurements to study effects of deiodinase subtypes. To our knowledge, this is the first IPD meta-analysis studying the association between thyroid hormone status and renal function. Findings from our study are strengthened by the large population size, its global representativeness and availability of IPD, making it possible to standardize definitions, statistical models and outcomes. Several limitations need to be discussed. First, methodology of creatinine measurements was not similar across studies. Since differences in renal function were calculated between groups and individuals on a study level, this cannot have resulted in systematic error. For some studies, different assays were used between the visits in the longitudinal analyses, which may have resulted in dilution of the results to the null. For example, in the HUNT study and Health ABC, different assays were used at baseline and during follow-up. However, sensitivity analyses excluding results from the HUNT study and Health ABC did not change our findings. Also, eGFR is an approximation of renal function. Determining measured GFR would benefit classification of individuals in their outcome. Nevertheless, estimation equations have been shown to be accurate for following changes in GFR over time [41]. Secondly, our study was not specifically designed to study the impact of overt hypothyroidism on renal function. Only a minority of individuals in our study had TSH levels >20 mIU/L. Also, thyroid hormone usage was more prevalent in the hypothyroid and subclinically hypothyroid groups and could have prevented its downstream effects on renal function. However, adjustment for thyroid medication revealed no differences in findings suggests that overt hypothyroidism is not associated with an additional decline in renal function over time. Potential confounding effects of other drugs such as amiodarone, glucocorticoids and lithium could not be determined because these parameters were not available in most cohorts. Since relatively few individuals in the general population use these medications, effects on parameters are likely small. Finally, censoring due to events of death could have caused selection bias. Meta-regression analyses did, however, not reveal an association between the proportion of mortality and the effect estimate in the studies included. Also, estimation equations have been shown to be accurate for assessing GFR slopes over time and its determinants [42, 43], which supports the legitimacy of studying thyroid dysfunction as a risk factor for changes in renal function in these cohorts. Overall, we found that, cross-sectionally, low thyroid function was associated with lower eGFR values as compared with euthyroid subjects. During follow-up, subjects with low thyroid function did not have a more pronounced decline in renal function over time than euthyroid subjects. We conclude that low thyroid function, and especially subclinical hypothyroidism, is not associated with deterioration in renal function and speculate that cross-sectional findings may be explained by renal dysfunction causing thyroid hormone alterations. Further studies should shed light on the link between thyroid and renal function and possible differences among causes of thyroid disease. SUPPLEMENTARY DATA Supplementary data are available at ndt online. FUNDING The Busselton Health Study had no financial support to disclose. The Cardiovascular Health Study (CHS) was supported by contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083 and N01HC85086, and grants U01HL080295 and U01HL130114 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided by R01AG023629 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at CHS-NHLBI.org. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The European Prospective Investigation of Cancer (EPIC)-Norfolk study was supported by research grants from the Medical Research Council UK and Cancer Research UK. The Health, Aging and Body Composition (Health ABC) study was supported by NIA Contracts N01-AG-6-2101; N01-AG-6-2103; N01-AG-6-2106; NIA grant R01-AG028050 and NINR grant R01-NR012459. This research was funded in part by the Intramural Research Program at the NIA. The InCHIANTI study was supported as a target project ICS 110.1|RS97.71 by the Italian Ministry of Health, and in part by the US National Institute on Aging, contracts 263-MD-9164-13 and 263-MD-821336. The Nord-Trøndelag Health (HUNT) study is a collaborative effort of HUNT Research Center (Faculty of Medicine and Health Sciences, NTNU, Norwegian University of Science and Technology), the Norwegian Institute of Public Health, Central Norway Health Authority and the Nord-Trøndelag County Council. Thyroid function testing in the HUNT Study was financially supported by WallacOy (Turku, Finland). The Leiden 85-plus study was partly funded by the Dutch Ministry of Health, Welfare and Sports. The original PROSPER study was supported by an unrestricted, investigator-initiated grant from Bristol-Myers Squibb. The Rotterdam Study was funded by the following: Erasmus MC and Erasmus University, Rotterdam, the Netherlands; the Netherlands Organisation for Scientific Research (NWO); the Netherlands Organisation for the Health Research and Development (ZonMw); the Research Institute for Diseases in the Elderly (RIDE); the Ministry of Education, Culture and Science; the Dutch Ministry for Health, Welfare and Sports; the European Commission (DG XII); and the Municipality of Rotterdam. The Radiation Effects Research Foundation (RERF), Hiroshima and Nagasaki, Japan, is a public interest foundation funded by the Japanese Ministry of Health, Labour and Welfare (MHLW) and the US Department of Energy (DOE). This publication was supported by RERF Research Protocol A5–13. The views of the authors do not necessarily reflect those of the two governments. SHIP is part of the Research Network of Community Medicine at the University Medicine Greifswald, Germany (www.community-medicine.de), which is funded by the German Federal State of Mecklenburg–West Pomerania. The BELFRAIL study is funded by an unconditional grant from the Fondation Louvain. The Fondation Louvain is the support unit of the Université Catholique de Louvain in charge of developing education and research projects of the university by collecting gifts from corporate, foundations and alumni. The Brazilian thyroid study was supported by an unrestricted grant from São Paulo State Research Foundation (Fundação de Amparo a Pesquisa do Estado de São Paulo) Grant 6/59737-9 (to R.M.B.M.). The Prevention of Renal and Vascular End-Stage Disease (PREVEND) study has been made possible by grants from the Dutch Kidney Foundation. The work from N.R. was supported by grants from the Swiss National Science Foundation (SNSF 320030-150025 and 320030-172676 both to N.R.). CONFLICT OF INTEREST STATEMENT The sponsor had no role in the design and conduct of the study; in the collection, analysis and interpretation of the data; or in the preparation, review or approval of the manuscript. The authors have no conflicts to report. REFERENCES 1 Stenvinkel P. Chronic kidney disease: a public health priority and harbinger of premature cardiovascular disease. J Intern Med  2010; 268: 456– 467 Google Scholar CrossRef Search ADS PubMed  2 Foley RN, Murray AM, Li S et al.   Chronic kidney disease and the risk for cardiovascular disease, renal replacement, and death in the United States Medicare population, 1998 to 1999. J Am Soc Nephrol  2005; 16: 489– 495 Google Scholar CrossRef Search ADS PubMed  3 de Jager DJ, Grootendorst DC, Jager KJ et al.   Cardiovascular and noncardiovascular mortality among patients starting dialysis. JAMA  2009; 302: 1782– 1789 Google Scholar CrossRef Search ADS PubMed  4 Song SH, Kwak IS, Lee DW et al.   The prevalence of low triiodothyronine according to the stage of chronic kidney disease in subjects with a normal thyroid-stimulating hormone. Nephrol Dial Transplant  2009; 24: 1534– 1538 Google Scholar CrossRef Search ADS PubMed  5 Kaptein EM, LoPresti JS, Kaptein MJ. Is an isolated TSH elevation in chronic nonthyroidal illness “subclinical hypothyroidism”? J Clin Endocrinol Metab  2014; 99: 4015– 4026 Google Scholar CrossRef Search ADS PubMed  6 Anderson JLC, Gruppen EG, van Tienhoven-Wind L et al.   Glomerular filtration rate is associated with free triiodothyronine in euthyroid subjects: comparison between various equations to estimate renal function and creatinine clearance. Eur J Intern Med  2017; 48: 94– 99 Google Scholar CrossRef Search ADS PubMed  7 Meuwese CL, Dekkers OM, Stenvinkel P et al.   Nonthyroidal illness and the cardiorenal syndrome. Nat Rev Nephrol  2013; 9: 599– 609 Google Scholar CrossRef Search ADS PubMed  8 Montenegro J, Gonzalez O, Saracho R et al.   Changes in renal function in primary hypothyroidism. Am J Kidney Dis  1996; 27: 195– 198 Google Scholar CrossRef Search ADS PubMed  9 Kreisman SH, Hennessey JV. Consistent reversible elevations of serum creatinine levels in severe hypothyroidism. Arch Intern Med  1999; 159: 79– 82 Google Scholar CrossRef Search ADS PubMed  10 Gencer B, Collet TH, Virgini V et al.   Subclinical thyroid dysfunction and cardiovascular outcomes among prospective cohort studies. Endocr Metab Immune Disord Drug Targets  2013; 13: 4– 12 Google Scholar CrossRef Search ADS PubMed  11 Rodondi N, den Elzen WP, Bauer DC et al.   Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA  2010; 304: 1365– 1374 Google Scholar CrossRef Search ADS PubMed  12 Shin DH, Lee MJ, Kim SJ et al.   Preservation of renal function by thyroid hormone replacement therapy in chronic kidney disease patients with subclinical hypothyroidism. J Clin Endocrinol Metab  2012; 97: 2732– 2740 Google Scholar CrossRef Search ADS PubMed  13 Shin DH, Lee MJ, Lee HS et al.   Thyroid hormone replacement therapy attenuates the decline of renal function in chronic kidney disease patients with subclinical hypothyroidism. Thyroid  2013; 23: 654– 661 Google Scholar CrossRef Search ADS PubMed  14 Meuwese CL, Gussekloo J, de Craen AJ et al.   Thyroid status and renal function in older persons in the general population. J Clin Endocrinol Metab  2014; 99: 2689– 2696 Google Scholar CrossRef Search ADS PubMed  15 Cappola AR, Fried LP, Arnold AM et al.   Thyroid status, cardiovascular risk, and mortality in older adults. JAMA  2006; 295: 1033– 1041 Google Scholar CrossRef Search ADS PubMed  16 Rodondi N, Newman AB, Vittinghoff E et al.   Subclinical hypothyroidism and the risk of heart failure, other cardiovascular events, and death. Arch Intern Med  2005; 165: 2460– 2466 Google Scholar CrossRef Search ADS PubMed  17 Boekholdt SM, Titan SM, Wiersinga WM et al.   Initial thyroid status and cardiovascular risk factors: the EPIC-Norfolk prospective population study. Clin Endocrinol (Oxf)  2010; 72: 404– 410 Google Scholar CrossRef Search ADS PubMed  18 Asvold BO, Bjoro T, Nilsen TI et al.   Thyrotropin levels and risk of fatal coronary heart disease: the HUNT study. Arch Intern Med  2008; 168: 855– 860 Google Scholar CrossRef Search ADS PubMed  19 Gussekloo J, van Exel E, de Craen AJ et al.   Thyroid status, disability and cognitive function, and survival in old age. JAMA  2004; 292: 2591– 2599 Google Scholar CrossRef Search ADS PubMed  20 Iervasi G, Molinaro S, Landi P et al.   Association between increased mortality and mild thyroid dysfunction in cardiac patients. Arch Intern Med  2007; 167: 1526– 1532 Google Scholar CrossRef Search ADS PubMed  21 Sgarbi JA, Matsumura LK, Kasamatsu TS et al.   Subclinical thyroid dysfunctions are independent risk factors for mortality in a 7.5-year follow-up: the Japanese-Brazilian thyroid study. Eur J Endocrinol  2010; 162: 569– 577 Google Scholar CrossRef Search ADS PubMed  22 Walsh JP, Bremner AP, Bulsara MK et al.   Subclinical thyroid dysfunction as a risk factor for cardiovascular disease. Arch Intern Med  2005; 165: 2467– 2472 Google Scholar CrossRef Search ADS PubMed  23 Imaizumi M, Akahoshi M, Ichimaru S et al.   Risk for ischemic heart disease and all-cause mortality in subclinical hypothyroidism. J Clin Endocrinol Metab  2004; 89: 3365– 3370 Google Scholar CrossRef Search ADS PubMed  24 Iacoviello M, Guida P, Guastamacchia E et al.   Prognostic role of sub-clinical hypothyroidism in chronic heart failure outpatients. Curr Pharm Des  2008; 14: 2686– 2692 Google Scholar CrossRef Search ADS PubMed  25 Vaes B, Pasquet A, Wallemacq P et al.   The BELFRAIL (BFC80+) study: a population-based prospective cohort study of the very elderly in Belgium. BMC Geriatr  2010; 10: 39 Google Scholar CrossRef Search ADS PubMed  26 Nanchen D, Gussekloo J, Westendorp RG et al.   Subclinical thyroid dysfunction and the risk of heart failure in older persons at high cardiovascular risk. J Clin Endocrinol Metab  2012; 97: 852– 861 Google Scholar CrossRef Search ADS PubMed  27 Ittermann T, Haring R, Sauer S et al.   Decreased serum TSH levels are not associated with mortality in the adult northeast German population. Eur J Endocrinol  2010; 162: 579– 585 Google Scholar CrossRef Search ADS PubMed  28 van der Deure WM, Peeters RP, Uitterlinden AG et al.   Impact of thyroid function and polymorphisms in the type 2 deiodinase on blood pressure: the Rotterdam Study and the Rotterdam Scan Study. Clin Endocrinol (Oxf)  2009; 71: 137– 144 Google Scholar CrossRef Search ADS PubMed  29 Ceresini G, Lauretani F, Maggio M et al.   Thyroid function abnormalities and cognitive impairment in elderly people: results of the Invecchiare in Chianti study. J Am Geriatr Soc  2009; 57: 89– 93 Google Scholar CrossRef Search ADS PubMed  30 Deetman PE, Bakker SJ, Kwakernaak AJ et al.   The relationship of the anti-oxidant bilirubin with free thyroxine is modified by insulin resistance in euthyroid subjects. PLoS One  2014; 9: e90886 Google Scholar CrossRef Search ADS PubMed  31 Hallan S, Astor B, Lydersen S. Estimating glomerular filtration rate in the general population: the second Health Survey of Nord-Trondelag (HUNT II). Nephrol Dial Transplant  2006; 21: 1525– 1533 Google Scholar CrossRef Search ADS PubMed  32 Levey AS, Bosch JP, Lewis JB et al.   A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med  1999; 130: 461– 470 Google Scholar CrossRef Search ADS PubMed  33 Levey AS, Stevens LA, Schmid CH et al.   A new equation to estimate glomerular filtration rate. Ann Intern Med  2009; 150: 604– 612 Google Scholar CrossRef Search ADS PubMed  34 Ferguson MA, Waikar SS. Established and emerging markers of kidney function. Clin Chem  2012; 58: 680– 689 Google Scholar CrossRef Search ADS PubMed  35 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials  1986; 7: 177– 188 Google Scholar CrossRef Search ADS PubMed  36 Asvold BO, Bjoro T, Vatten LJ. Association of thyroid function with estimated glomerular filtration rate in a population-based study. The HUNT Study. Eur J Endocrinol  2011; 164: 101– 105 Google Scholar CrossRef Search ADS PubMed  37 Rhee CM, Kalantar-Zadeh K, Streja E et al.   The relationship between thyroid function and estimated glomerular filtration rate in patients with chronic kidney disease. Nephrol Dial Transplant  2015; 30: 282– 287 Google Scholar CrossRef Search ADS PubMed  38 Mandelli S, Riva E, Tettamanti M et al.   Mortality prediction in the oldest old with five different equations to estimate glomerular filtration rate: the health and anemia population-based study. PLoS One  2015; 10: e0136039 Google Scholar CrossRef Search ADS PubMed  39 Kaptein EM. Thyroid hormone metabolism and thyroid diseases in chronic renal failure. Endocr Rev  1996; 17: 45– 63 Google Scholar CrossRef Search ADS PubMed  40 Warner MH, Beckett GJ. Mechanisms behind the non-thyroidal illness syndrome: an update. J Endocrinol  2010; 205: 1– 13 Google Scholar CrossRef Search ADS PubMed  41 Wang X, Lewis J, Appel L et al.   Validation of creatinine-based estimates of GFR when evaluating risk factors in longitudinal studies of kidney disease. J Am Soc Nephrol  2006; 17: 2900– 2909 Google Scholar CrossRef Search ADS PubMed  42 Hallan SI, Ritz E, Lydersen S et al.   Combining GFR and albuminuria to classify CKD improves prediction of ESRD. J Am Soc Nephrol  2009; 20: 1069– 1077 Google Scholar CrossRef Search ADS PubMed  43 Verhave JC, Gansevoort RT, Hillege HL et al.   An elevated urinary albumin excretion predicts de novo development of renal function impairment in the general population. Kidney Int Suppl  2004; 92: S18– S21 Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nephrology Dialysis Transplantation Oxford University Press

Loading next page...
 
/lp/ou_press/low-thyroid-function-is-not-associated-with-an-accelerated-R4UGVm0nds
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
ISSN
0931-0509
eISSN
1460-2385
D.O.I.
10.1093/ndt/gfy071
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Chronic kidney disease (CKD) is frequently accompanied by thyroid hormone dysfunction. It is currently unclear whether these alterations are the cause or consequence of CKD. This study aimed at studying the effect of thyroid hormone alterations on renal function in cross-sectional and longitudinal analyses in individuals from all adult age groups. Methods Individual participant data (IPD) from 16 independent cohorts having measured thyroid stimulating hormone, free thyroxine levels and creatinine levels were included. Thyroid hormone status was defined using clinical cut-off values. Estimated glomerular filtration rates (eGFR) were calculated by means of the four-variable Modification of Diet in Renal Disease (MDRD) formula. For this IPD meta-analysis, eGFR at baseline and eGFR change during follow-up were computed by fitting linear regression models and linear mixed models in each cohort separately. Effect estimates were pooled using random effects models. Results A total of 72 856 individuals from 16 different cohorts were included. At baseline, individuals with overt hypothyroidism (n = 704) and subclinical hypothyroidism (n = 3356) had a average (95% confidence interval) −4.07 (−6.37 to −1.78) and −2.40 (−3.78 to −1.02) mL/min/1.73 m2 lower eGFR as compared with euthyroid subjects (n = 66 542). In (subclinical) hyperthyroid subjects (n = 2254), average eGFR was 3.01 (1.50–4.52) mL/min/1.73 m2 higher. During 329 713 patient years of follow-up, eGFR did not decline more rapidly in individuals with low thyroid function compared with individuals with normal thyroid function. Conclusions Low thyroid function is not associated with a deterioration of renal function. The cross-sectional association may be explained by renal dysfunction causing thyroid hormone alterations. chronic renal failure, CKD, creatinine clearance, epidemiology, thyroid function INTRODUCTION The prevalence of chronic kidney disease (CKD) is globally increasing, reaching endemic levels [1]. Since this growth is accompanied by a substantial increase in cardiovascular morbidity and mortality [2, 3], prevention of CKD and its secondary complications are of increasing importance. To date, however, aggressive management of known risk factors such as blood pressure, albuminuria and glucose control has not resulted in a clear reduction of this trend. These observations stress the need for further studies on other risk factors being amenable for treatment. In cross-sectional studies, lower renal function is accompanied by reductions in free thyroxine (fT4) and triiodothyronine (T3) and an elevation in serum thyroid stimulating hormone (TSH) levels [4–6]. This finding can be interpreted in two ways: first, CKD (similar to other chronic illnesses) induces a systemic lowering of the hypothalamic–pituitary–thyroid (HPT) axis, known as ‘non-thyroidal illness’ [7]. Alternatively, primary hypothyroidism could be the cause of a reduction in renal function. Indeed, studies in patients with severe primary hypothyroidism show a consistent reduction in renal function, which resolves after initiation of thyroid hormone supplementation [8, 9]. Although large-scale observational studies show clear associations between subclinical hypothyroidism and an increased risk for heart failure [10], coronary heart disease and mortality [11] findings are not consistent with the association between subclinical hypothyroidism and lower renal function [12–14]. In an observational study in patients with CKD Stages 2–4 and subclinical hypothyroidism, subjects not being prescribed thyroid hormone treatment showed a more rapid decline of renal function as compared with patients who received thyroid hormone supplementation [12, 13]. However, no association between thyroid hormone status and a decline in renal function was observed in a population-based study of the oldest old [14]. In light of these conflicting findings, this study sets out to evaluate the association between thyroid hormone status and renal function cross-sectionally and longitudinally by performing an individual patient data (IPD) meta-analyses on data from 16 independent cohorts participating in the Thyroid Studies Collaboration. MATERIALS AND METHODS Data from 16 different cohorts (four from the Netherlands, three from Italy, two from USA, two from Japan (and partly Brazil), and one from Germany, Norway, Belgium, Australia and Ireland), providing measures of thyroid and renal function were used for our analyses [11]. Nine of these cohorts [15–23] were also included in an earlier study evaluating the association between thyroid function and cardiovascular mortality [11]. Details of these cohorts have been described previously. In addition to these nine cohorts, seven cohorts [24–30] with data on thyroid and renal function were added to the collaboration. Six out of these seven cohorts were population-based studies; two comprising individuals from all age categories [27, 30], two having included those with an average age ∼70 years [28, 29] and two other studies included specifically the oldest [25, 26]. The seventh study was comprised of patients with chronic heart failure [24]. Thyroid function Thyroid function tests were measured at baseline in each cohort. We used common definitions to define thyroid hormone groups by using cohort-specific cut-off values which are summarized in Supplementary data, Appendix S1: (i) overt hypothyroidism was defined as elevated TSH levels in combination with reduced fT4 levels; (ii) subclinical hypothyroidism was defined as an elevated serum TSH level with a normal fT4 concentration; (iii) subjects were categorized in the euthyroid group when having TSH levels within the specific reference range; and (iv) those with lowered TSH levels with or without elevated fT4 levels were categorized as (subclinically) hyperthyroid. Subjects with subclinical hyperthyroidism and overt hyperthyroidism were combined because of low numbers in each group. Renal function Creatinine levels were measured according to each cohort’s protocol. Differences exist in their methodology; five cohorts utilized colorimetric assessments [15–17, 22, 25], five cohorts utilized the traditional Jaffé method [19, 21, 26, 27, 29] and the newer enzymatic method was applied in four cohorts [20, 23, 28, 30]. In the Bari cohort [24], creatinine levels were measured with both the Jaffé and colorimetric method. In the Nord-Trøndelag Health (HUNT) study [18], the baseline examination (HUNT2) was performed using the Jaffé method and subsequently adjusted by means of a validated calibration formula [31]. The alkaline picrate methodology was used for the follow-up examination (HUNT3) . Estimated glomerular filtrations rate (eGFR) was assessed by means of the four-variable Modification of Diet in Renal Disease (MDRD) formula [32]. For sensitivity analyses, eGFR was calculated on basis of the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula [33]. The MDRD formula was used as primary outcome instead of the CKD-EPI because not all creatinine measurements were based on traceable isotope dilution mass spectrometry [34]. Statistical analyses Baseline characteristics for each cohort are presented as means with SDs or numbers with percentages (%), as appropriate. To analyse the association between thyroid hormone status and renal function, a two-stage IPD meta-analysis was used as previously specified [11]. First, effect estimates were calculated at a cohort level. Thereafter, they were pooled at a meta-analysis level. For the cross-sectional associations between thyroid hormone status and renal function, linear regression analyses were fitted. Thyroid hormone groups were entered as categorical variables with the euthyroid group serving as the reference. Effect estimates (betas) represented the difference in eGFR (mL/min/1.73 m2) at baseline for the specific thyroid hormone group with respect to the euthyroid group. The same concept was applied to TSH and fT4 groups, again those with normal levels serving as the reference category. TSH and fT4 were also entered as continuous variables in which effect estimates illustrated an increase in eGFR per 1 mIU/L and per 1 pmol/L increase in serum TSH and fT4 levels, respectively. For the longitudinal analyses, examining the association between thyroid hormone status at baseline and the change in renal function over time, linear mixed models were fitted in each cohort separately. Because of the large variability in the number and timing of measurement points within cohorts and between cohorts, we chose to adopt random effects models. An average change in eGFR (mL/min/1.73 m2/year) for each cohort was calculated and presented in a figure. Slopes were not pooled because of large heterogeneity in effect estimates and participant characteristics. As for the linear regression analyses, thyroid hormone groups were entered as categorical variables. In addition, an interaction term of thyroid hormone group and time was included to allow for dependence of the slope on thyroid hormone status. Effect estimates obtained from these models indicated the additional change in eGFR per year as compared with the change in the euthyroid group. All models were adjusted for age, sex, cardiovascular disease and when available, for thyroid hormone supplementation and/or anti-thyroid medication. In sensitivity analyses, models were rerun also adjusting for diabetes mellitus, if available. Outcomes obtained from linear regression analyses (cross-sectionally) and from linear mixed models (longitudinally) were pooled by means of random effects models assuming the variance model as proposed by DerSimonian and Laird [35]. Sensitivity analyses were performed by rerunning all previous models in subgroups of sex and age. Another sensitivity analysis was performed excluding the HUNT study because creatinine measurements at baseline and follow-up were performed with different assays. The same was done for the Health, Aging and Body Composition (Health ABC) study. Also, sensitivity analyses were done by excluding all cohorts with positive changes in eGFR over time [22, 26, 27, 29]. To further examine the potential of selection/publication bias, funnel plots were created. Bubble plots were created plotting the effect size against mortality rates in each cohort. For differences, a 95% confidence interval (CI) not including zero was considered to indicate statistical significance. For all other tests, a P-value <0.05 was adopted as cut-off. Stata 12.1 (StataCorp LP, Texas, USA ) was used to perform all analyses. Figures were created using Stata and Prism 5.02 (GraphPad Software Inc., La Jolla, CA, USA; 1992). RESULTS This study included data from 16 cohorts, comprising a total of 72 856 individuals of whom 704 were hypothyroid, 3356 subclinically hypothyroid, 66 542 euthyroid and 2254 (subclinically) hyperthyroid. Baseline characteristics of the different cohorts are presented in Table 1. As illustrated, the average age at baseline ranged from 49 to 85 years and the proportion of individuals with pre-existing cardiovascular disease from 1.9% to 100%. Within the different cohorts, 0–9.9% of subjects were prescribed thyroid hormone replacement therapy and 0–4.7% used antithyroid medication. Usage of thyroid hormone supplementation was more common in the subclinical hypothyroid and hypothyroid groups, whereas more individuals in the hyperthyroid group used antithyroid medication (Supplementary data, Appendix S2). Table 1. Baseline characteristics of the different cohorts Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  a Combined thyroid supplementation/antithyroid medication. CHS, Cardiovascular Health Study; EPIC, European Prospective Investigation into Cancer and Nutrition Study; SHIP, Study of Health in Pomerania; InCHIANTI, Invecchiare in Chianti; PREVEND, Prevention of Renal and Vascular End-Stage Disease; na, not available; subcl, subclinical; hypo, hypothyroidism. Table 1. Baseline characteristics of the different cohorts Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  Cohort  Country  Total no. of participants/ mean years of follow-up per individual/no. of serum measurements per individual  Percentage of men  Average age (SD), years  Part with CVD (%)  Part with hypo/subcl hypo/(subcl) hyper (%)  Part using thyroxine/antithyroid medication at baseline (%)  CHS [15]  USA  3112/6.7/4  40.0  72.6 (5.6)  1.9  1.2/15.9/0.4  0/0  Health ABC study [16]  USA  2776/4.5/3  48.9  74.7 (2.9)  30.3  0.9/4.5/0.4  9.9/0  EPIC study [17]  UK  9869  43.5  59.0 (9.2)  4.6  1.8/5.6/4.5  na  Bari study [24]  Italy  338/2.4/3  76.9  64.3 (13.0)  100  0/12.1/3.3  5.0/1.8  HUNT study [18]  Norway  33927/11.2/2  31.4  58.6 (13.4)  10.6  0.6/4.1/1.7  5.0/0.2  BELFRAIL study [25]  Belgium  542/1.7/2  37.3  84.8 (0.4)  58.7  3.9/2.2/4.0  9.7/0.0  Leiden 85-plus study [19]  The Netherlands  558/3.9/6  33.9  85  48.6  7.2/6.1/5.3  2.9/0.7  Pisa study [20]  Italy  2260  65.3  65.8 (13.0)  98.5  0.1/5.2/6.3  0  PROSPER study [26]  The Netherlands  5794/0.3/2  48.3  75.3 (3.3)  43.9  0.6/3.7/3.4  4.4/0.1  SHIP study [27]  Germany  4236/9.3/3  49.2  49.7 (16.3)  6.1  0.3/2.8/8.5  6.3a  Busselton Health study [22]  Australia  832/13.0/2  46.8  52.8 (10.3)  5.8  0.9/4.9/3.7  1.1/0.0  Japanese-Brazilian Thyroid study [20]  Japan/Brazil  1110  46.8  56.5 (12.5)  14.1  1.0/8.9/11.2  0/0  RERF [23]  Japan  1730/7.5/7  32.9  69.0 (8.8)  na  3.9/6.6/3.2  3.8/0.2  Rotterdam study [28]  The Netherlands  1875/4.7/2  38.3  68.8 (7.5)  27.8  2.0/3.1/6.7  2.5  InCHIANTI [29]  Italy  1209/8.9/4  43.5  69.0 (0.4)  12.4  0.8/2.7/10.2  2.1/0.7  PREVEND [30]  The Netherlands  2688/3.4/6  48.4  48.5 (12.6)  9.1  1.0/1.5/2.7  na  a Combined thyroid supplementation/antithyroid medication. CHS, Cardiovascular Health Study; EPIC, European Prospective Investigation into Cancer and Nutrition Study; SHIP, Study of Health in Pomerania; InCHIANTI, Invecchiare in Chianti; PREVEND, Prevention of Renal and Vascular End-Stage Disease; na, not available; subcl, subclinical; hypo, hypothyroidism. Cross-sectional analyses In Figure 1, the average eGFR per cohort is shown with mean (SD) values ranging from 59.0 (14.4) mL/min/1.73 m2 in the Leiden 85-plus study to 102.6 (26.7) mL/min/1.73 m2 in Radiation Effects Research Foundation (RERF). Figure 2 details forest plots presenting the differences in eGFR at baseline between the hypothyroid, subclinical hypothyroid and subclinical hyperthyroid group versus euthyroid group for each cohort separately. Pooled estimates show that eGFR was on average −4.07 (95% CI: −6.37 to −1.78) mL/min/1.73 m2 lower in the hypothyroid and −2.40 (−3.78 to −1.02) mL/min/1.73 m2 lower in the subclinical hypothyroid group as compared with the euthyroid group. Conversely, average eGFR was 3.01 (1.50–4.52) mL/min/1.73 m2 higher in (subclinical) hyperthyroid subjects. FIGURE 1: View largeDownload slide Mean (SD) eGFR (mL/min/1.73 m2) at baseline in the different cohorts. FIGURE 1: View largeDownload slide Mean (SD) eGFR (mL/min/1.73 m2) at baseline in the different cohorts. FIGURE 2: View largeDownload slide Forest plots providing the pooled differences in eGFR (mL/min/1.73 m2) at baseline in each thyroid group as compared with the euthyroid group. FIGURE 2: View largeDownload slide Forest plots providing the pooled differences in eGFR (mL/min/1.73 m2) at baseline in each thyroid group as compared with the euthyroid group. FIGURE 2: View largeDownload slide Continued. FIGURE 2: View largeDownload slide Continued. Longitudinal analyses Of the 16 cohorts, 13 contributed a total of 113,670 measurements during 329 713 patient-years of follow-up. Figure 3 depicts adjusted average annual changes in eGFR (mL/min/1.73 m2/year) per cohort, showing a range in annual change from −1.43 (0.06) to + 8.98 (1.05) mL/min/1.73 m2/year. Figure 4 demonstrates the pooled differences in eGFR change per year within the different thyroid hormone groups as compared with the euthyroid group. The change in eGFR was 0.35 (0.07–0.64) mL/min/1.73 m2 per year higher in the overt hypothyroid compared with the euthyroid group. No significant differences in eGFR change over time were noted in the other thyroid hormone groups as compared with euthyroid subjects. FIGURE 3: View largeDownload slide Mean (SD) changes in eGFR (mL/min/1.73 m2/year) in the different cohorts. FIGURE 3: View largeDownload slide Mean (SD) changes in eGFR (mL/min/1.73 m2/year) in the different cohorts. FIGURE 4: View largeDownload slide Forest plots providing the pooled additional changes in eGFR (mL/min/1.73 m2/year) per year in each thyroid group as compared with the euthyroid group. n = number of observations. FIGURE 4: View largeDownload slide Forest plots providing the pooled additional changes in eGFR (mL/min/1.73 m2/year) per year in each thyroid group as compared with the euthyroid group. n = number of observations. FIGURE 4: View largeDownload slide Continued. FIGURE 4: View largeDownload slide Continued. All analyses were repeated cross-sectionally and longitudinally in subgroups of sex (Supplementary data, Appendices S3 and S5) and age (<50, 50–65, 65–80 and >80 years) (Supplementary data, Appendices S4 and S6) rendering no differences in results. When models were adjusted for diabetes mellitus, results were not substantially different (Supplementary data, Appendices S7 and S8). Using the CKD-EPI formula instead of the four-variable MDRD formula, findings were comparable (data not shown). When pooling was repeated excluding the HUNT and Health ABC studies, similar results were found. The same held for the analyses excluding cohorts with positive changes in eGFR over time. Finally, analyses were repeated in individuals not using thyroid hormone supplementation or antithyroid medication, with no effect on the findings (data not shown). To examine the possibility of selection bias as a consequence of death, study-specific effect estimates were regressed on mortality incidence rates for each study, showing no statistically significant association (data not shown). Finally, a funnel plot did not show evidence of publication bias (data not shown). DISCUSSION In this IPD meta-analysis comprising data from 72 856 individuals out of 16 independent cohorts, we found a positive cross-sectional association between thyroid function and renal function in which those with low thyroid function had lower eGFR values as compared with euthyroid and (subclinical) hyperthyroid subjects. During follow-up, low thyroid function was not associated with an additional decline in renal function as compared with the euthyroid group. The presence of a cross-sectional association between low thyroid function and renal dysfunction aligns with findings in several previous cohorts [14, 36, 37], two of which were also included in this meta-analysis [14, 36]. Also compatible with our previous report, low thyroid function did not associate with an additional decline in renal function versus a euthyroid state [14]. Rather, we observed a relative increase in renal function in subjects with overt hypothyroidism as compared with individuals with thyroid hormone concentrations within the reference range. Three potential clarifications should be considered: first, whereas those with hypothyroidism have a lower eGFR at baseline, the observation of an increase in eGFR values over time relative to the euthyroid group may be explained by the concept of ‘regression to the mean’. Regression to the mean implies that when a variable has extreme values at a certain measurement, a second measurement will tend, by chance, to show a value closer to the true mean. In case of an extremely high value, the second measurement tends to be lower and vice versa, in case of an extremely low value, the second measurement will tend to portrait a higher value. Secondly, as overt hypothyroidism is generally considered an indication for thyroid hormone supplementation, the relative rise in eGFR over time could be the resultant of treatment rather than due to hypothyroidism itself. Yet, sensitivity analyses excluding subjects on thyroid medication did not reveal differential findings. Nevertheless, changes in thyroid status and medication during follow-up could have translated into different outcomes. Finally, patients with hypothyroidism could be more inactive resulting in a lower muscle mass, thereby lowering creatinine levels. We did not observe a difference in eGFR change over time in the subclinical hypothyroid group as compared with euthyroid subjects. Because subclinical hypothyroidism is not a strict indication for thyroid hormone supplementation, thyroid hormone supplementation likely did not play an interfering role. Results of a previous non-randomized study suggest that thyroxine supplementation preserves renal function over time in patients with CKD Stages 2–4 [12]. Findings from that study are, however, hampered by several limitations. First, in the absence of a randomized design and/or appropriate adjustment, confounding by indication may have imposed systematic error. Patient characteristics and physician preferences likely influenced the decision to initiate treatment. The non-treatment group indeed seemed overall less healthy than the treatment group. Secondly, 49 out of 358 individuals were excluded from the analyses because of a follow-up duration <12 months. This loss to follow-up may have been dependent on treatment status and outcome, and as a result, have introduced selection bias. Therefore, in addition to previous literature, current findings do not support a causal relationship between subclinical hypothyroidism and a decline in renal function over time. It is of interest to speculate why eGFR increased over time in cohorts with an average higher age [25, 26]. One of the explanations may be that those individuals with more rapid declines in renal function died sooner. Alternatively, as muscle metabolism and habitus change in old age, conventional equations for assessing eGFR are poorly validated in elderly individuals [38]. Nevertheless, in the Leiden 85-plus study, comprising individuals in similar age categories as PROSPER (Prospective Study of Pravastatin in the Elderly at Risk) and BELFRAIL (the Belgian cohort of the Very Elderly), average annual change in eGFR was considerably lower. When repeating our analyses excluding cohorts with positive changes in eGFR, results did not change, leading us to believe that this paradoxical increase in eGFR would not have translated into bias. Given the absence of a longitudinal association, the concept of reversed causation (CKD causing thyroid hormone abnormalities) may explain the observed cross-sectional association between low thyroid and renal function in our study. CKD, and especially end-stage renal disease, is frequently accompanied by abnormal TSH, low triiodothyronine (fT3) and fT4 levels fitting the spectrum of so-called ‘non-thyroidal illness’ [39]. In the absence of primary disease in the HPT axis, its pathogenesis is multifactorial and occurs at multiple levels including peripheral deiodinase-dependent conversion defects and central alterations in thyroid hormone signalling [40]. It could be speculated that deiodinase defects in early phases prevail over central mechanisms, leading to a compensatory increase in TSH secretion. Further studies on this hypothesis could include (free) T3 measurements to study effects of deiodinase subtypes. To our knowledge, this is the first IPD meta-analysis studying the association between thyroid hormone status and renal function. Findings from our study are strengthened by the large population size, its global representativeness and availability of IPD, making it possible to standardize definitions, statistical models and outcomes. Several limitations need to be discussed. First, methodology of creatinine measurements was not similar across studies. Since differences in renal function were calculated between groups and individuals on a study level, this cannot have resulted in systematic error. For some studies, different assays were used between the visits in the longitudinal analyses, which may have resulted in dilution of the results to the null. For example, in the HUNT study and Health ABC, different assays were used at baseline and during follow-up. However, sensitivity analyses excluding results from the HUNT study and Health ABC did not change our findings. Also, eGFR is an approximation of renal function. Determining measured GFR would benefit classification of individuals in their outcome. Nevertheless, estimation equations have been shown to be accurate for following changes in GFR over time [41]. Secondly, our study was not specifically designed to study the impact of overt hypothyroidism on renal function. Only a minority of individuals in our study had TSH levels >20 mIU/L. Also, thyroid hormone usage was more prevalent in the hypothyroid and subclinically hypothyroid groups and could have prevented its downstream effects on renal function. However, adjustment for thyroid medication revealed no differences in findings suggests that overt hypothyroidism is not associated with an additional decline in renal function over time. Potential confounding effects of other drugs such as amiodarone, glucocorticoids and lithium could not be determined because these parameters were not available in most cohorts. Since relatively few individuals in the general population use these medications, effects on parameters are likely small. Finally, censoring due to events of death could have caused selection bias. Meta-regression analyses did, however, not reveal an association between the proportion of mortality and the effect estimate in the studies included. Also, estimation equations have been shown to be accurate for assessing GFR slopes over time and its determinants [42, 43], which supports the legitimacy of studying thyroid dysfunction as a risk factor for changes in renal function in these cohorts. Overall, we found that, cross-sectionally, low thyroid function was associated with lower eGFR values as compared with euthyroid subjects. During follow-up, subjects with low thyroid function did not have a more pronounced decline in renal function over time than euthyroid subjects. We conclude that low thyroid function, and especially subclinical hypothyroidism, is not associated with deterioration in renal function and speculate that cross-sectional findings may be explained by renal dysfunction causing thyroid hormone alterations. Further studies should shed light on the link between thyroid and renal function and possible differences among causes of thyroid disease. SUPPLEMENTARY DATA Supplementary data are available at ndt online. FUNDING The Busselton Health Study had no financial support to disclose. The Cardiovascular Health Study (CHS) was supported by contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083 and N01HC85086, and grants U01HL080295 and U01HL130114 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided by R01AG023629 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at CHS-NHLBI.org. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The European Prospective Investigation of Cancer (EPIC)-Norfolk study was supported by research grants from the Medical Research Council UK and Cancer Research UK. The Health, Aging and Body Composition (Health ABC) study was supported by NIA Contracts N01-AG-6-2101; N01-AG-6-2103; N01-AG-6-2106; NIA grant R01-AG028050 and NINR grant R01-NR012459. This research was funded in part by the Intramural Research Program at the NIA. The InCHIANTI study was supported as a target project ICS 110.1|RS97.71 by the Italian Ministry of Health, and in part by the US National Institute on Aging, contracts 263-MD-9164-13 and 263-MD-821336. The Nord-Trøndelag Health (HUNT) study is a collaborative effort of HUNT Research Center (Faculty of Medicine and Health Sciences, NTNU, Norwegian University of Science and Technology), the Norwegian Institute of Public Health, Central Norway Health Authority and the Nord-Trøndelag County Council. Thyroid function testing in the HUNT Study was financially supported by WallacOy (Turku, Finland). The Leiden 85-plus study was partly funded by the Dutch Ministry of Health, Welfare and Sports. The original PROSPER study was supported by an unrestricted, investigator-initiated grant from Bristol-Myers Squibb. The Rotterdam Study was funded by the following: Erasmus MC and Erasmus University, Rotterdam, the Netherlands; the Netherlands Organisation for Scientific Research (NWO); the Netherlands Organisation for the Health Research and Development (ZonMw); the Research Institute for Diseases in the Elderly (RIDE); the Ministry of Education, Culture and Science; the Dutch Ministry for Health, Welfare and Sports; the European Commission (DG XII); and the Municipality of Rotterdam. The Radiation Effects Research Foundation (RERF), Hiroshima and Nagasaki, Japan, is a public interest foundation funded by the Japanese Ministry of Health, Labour and Welfare (MHLW) and the US Department of Energy (DOE). This publication was supported by RERF Research Protocol A5–13. The views of the authors do not necessarily reflect those of the two governments. SHIP is part of the Research Network of Community Medicine at the University Medicine Greifswald, Germany (www.community-medicine.de), which is funded by the German Federal State of Mecklenburg–West Pomerania. The BELFRAIL study is funded by an unconditional grant from the Fondation Louvain. The Fondation Louvain is the support unit of the Université Catholique de Louvain in charge of developing education and research projects of the university by collecting gifts from corporate, foundations and alumni. The Brazilian thyroid study was supported by an unrestricted grant from São Paulo State Research Foundation (Fundação de Amparo a Pesquisa do Estado de São Paulo) Grant 6/59737-9 (to R.M.B.M.). The Prevention of Renal and Vascular End-Stage Disease (PREVEND) study has been made possible by grants from the Dutch Kidney Foundation. The work from N.R. was supported by grants from the Swiss National Science Foundation (SNSF 320030-150025 and 320030-172676 both to N.R.). CONFLICT OF INTEREST STATEMENT The sponsor had no role in the design and conduct of the study; in the collection, analysis and interpretation of the data; or in the preparation, review or approval of the manuscript. The authors have no conflicts to report. REFERENCES 1 Stenvinkel P. Chronic kidney disease: a public health priority and harbinger of premature cardiovascular disease. J Intern Med  2010; 268: 456– 467 Google Scholar CrossRef Search ADS PubMed  2 Foley RN, Murray AM, Li S et al.   Chronic kidney disease and the risk for cardiovascular disease, renal replacement, and death in the United States Medicare population, 1998 to 1999. J Am Soc Nephrol  2005; 16: 489– 495 Google Scholar CrossRef Search ADS PubMed  3 de Jager DJ, Grootendorst DC, Jager KJ et al.   Cardiovascular and noncardiovascular mortality among patients starting dialysis. JAMA  2009; 302: 1782– 1789 Google Scholar CrossRef Search ADS PubMed  4 Song SH, Kwak IS, Lee DW et al.   The prevalence of low triiodothyronine according to the stage of chronic kidney disease in subjects with a normal thyroid-stimulating hormone. Nephrol Dial Transplant  2009; 24: 1534– 1538 Google Scholar CrossRef Search ADS PubMed  5 Kaptein EM, LoPresti JS, Kaptein MJ. Is an isolated TSH elevation in chronic nonthyroidal illness “subclinical hypothyroidism”? J Clin Endocrinol Metab  2014; 99: 4015– 4026 Google Scholar CrossRef Search ADS PubMed  6 Anderson JLC, Gruppen EG, van Tienhoven-Wind L et al.   Glomerular filtration rate is associated with free triiodothyronine in euthyroid subjects: comparison between various equations to estimate renal function and creatinine clearance. Eur J Intern Med  2017; 48: 94– 99 Google Scholar CrossRef Search ADS PubMed  7 Meuwese CL, Dekkers OM, Stenvinkel P et al.   Nonthyroidal illness and the cardiorenal syndrome. Nat Rev Nephrol  2013; 9: 599– 609 Google Scholar CrossRef Search ADS PubMed  8 Montenegro J, Gonzalez O, Saracho R et al.   Changes in renal function in primary hypothyroidism. Am J Kidney Dis  1996; 27: 195– 198 Google Scholar CrossRef Search ADS PubMed  9 Kreisman SH, Hennessey JV. Consistent reversible elevations of serum creatinine levels in severe hypothyroidism. Arch Intern Med  1999; 159: 79– 82 Google Scholar CrossRef Search ADS PubMed  10 Gencer B, Collet TH, Virgini V et al.   Subclinical thyroid dysfunction and cardiovascular outcomes among prospective cohort studies. Endocr Metab Immune Disord Drug Targets  2013; 13: 4– 12 Google Scholar CrossRef Search ADS PubMed  11 Rodondi N, den Elzen WP, Bauer DC et al.   Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA  2010; 304: 1365– 1374 Google Scholar CrossRef Search ADS PubMed  12 Shin DH, Lee MJ, Kim SJ et al.   Preservation of renal function by thyroid hormone replacement therapy in chronic kidney disease patients with subclinical hypothyroidism. J Clin Endocrinol Metab  2012; 97: 2732– 2740 Google Scholar CrossRef Search ADS PubMed  13 Shin DH, Lee MJ, Lee HS et al.   Thyroid hormone replacement therapy attenuates the decline of renal function in chronic kidney disease patients with subclinical hypothyroidism. Thyroid  2013; 23: 654– 661 Google Scholar CrossRef Search ADS PubMed  14 Meuwese CL, Gussekloo J, de Craen AJ et al.   Thyroid status and renal function in older persons in the general population. J Clin Endocrinol Metab  2014; 99: 2689– 2696 Google Scholar CrossRef Search ADS PubMed  15 Cappola AR, Fried LP, Arnold AM et al.   Thyroid status, cardiovascular risk, and mortality in older adults. JAMA  2006; 295: 1033– 1041 Google Scholar CrossRef Search ADS PubMed  16 Rodondi N, Newman AB, Vittinghoff E et al.   Subclinical hypothyroidism and the risk of heart failure, other cardiovascular events, and death. Arch Intern Med  2005; 165: 2460– 2466 Google Scholar CrossRef Search ADS PubMed  17 Boekholdt SM, Titan SM, Wiersinga WM et al.   Initial thyroid status and cardiovascular risk factors: the EPIC-Norfolk prospective population study. Clin Endocrinol (Oxf)  2010; 72: 404– 410 Google Scholar CrossRef Search ADS PubMed  18 Asvold BO, Bjoro T, Nilsen TI et al.   Thyrotropin levels and risk of fatal coronary heart disease: the HUNT study. Arch Intern Med  2008; 168: 855– 860 Google Scholar CrossRef Search ADS PubMed  19 Gussekloo J, van Exel E, de Craen AJ et al.   Thyroid status, disability and cognitive function, and survival in old age. JAMA  2004; 292: 2591– 2599 Google Scholar CrossRef Search ADS PubMed  20 Iervasi G, Molinaro S, Landi P et al.   Association between increased mortality and mild thyroid dysfunction in cardiac patients. Arch Intern Med  2007; 167: 1526– 1532 Google Scholar CrossRef Search ADS PubMed  21 Sgarbi JA, Matsumura LK, Kasamatsu TS et al.   Subclinical thyroid dysfunctions are independent risk factors for mortality in a 7.5-year follow-up: the Japanese-Brazilian thyroid study. Eur J Endocrinol  2010; 162: 569– 577 Google Scholar CrossRef Search ADS PubMed  22 Walsh JP, Bremner AP, Bulsara MK et al.   Subclinical thyroid dysfunction as a risk factor for cardiovascular disease. Arch Intern Med  2005; 165: 2467– 2472 Google Scholar CrossRef Search ADS PubMed  23 Imaizumi M, Akahoshi M, Ichimaru S et al.   Risk for ischemic heart disease and all-cause mortality in subclinical hypothyroidism. J Clin Endocrinol Metab  2004; 89: 3365– 3370 Google Scholar CrossRef Search ADS PubMed  24 Iacoviello M, Guida P, Guastamacchia E et al.   Prognostic role of sub-clinical hypothyroidism in chronic heart failure outpatients. Curr Pharm Des  2008; 14: 2686– 2692 Google Scholar CrossRef Search ADS PubMed  25 Vaes B, Pasquet A, Wallemacq P et al.   The BELFRAIL (BFC80+) study: a population-based prospective cohort study of the very elderly in Belgium. BMC Geriatr  2010; 10: 39 Google Scholar CrossRef Search ADS PubMed  26 Nanchen D, Gussekloo J, Westendorp RG et al.   Subclinical thyroid dysfunction and the risk of heart failure in older persons at high cardiovascular risk. J Clin Endocrinol Metab  2012; 97: 852– 861 Google Scholar CrossRef Search ADS PubMed  27 Ittermann T, Haring R, Sauer S et al.   Decreased serum TSH levels are not associated with mortality in the adult northeast German population. Eur J Endocrinol  2010; 162: 579– 585 Google Scholar CrossRef Search ADS PubMed  28 van der Deure WM, Peeters RP, Uitterlinden AG et al.   Impact of thyroid function and polymorphisms in the type 2 deiodinase on blood pressure: the Rotterdam Study and the Rotterdam Scan Study. Clin Endocrinol (Oxf)  2009; 71: 137– 144 Google Scholar CrossRef Search ADS PubMed  29 Ceresini G, Lauretani F, Maggio M et al.   Thyroid function abnormalities and cognitive impairment in elderly people: results of the Invecchiare in Chianti study. J Am Geriatr Soc  2009; 57: 89– 93 Google Scholar CrossRef Search ADS PubMed  30 Deetman PE, Bakker SJ, Kwakernaak AJ et al.   The relationship of the anti-oxidant bilirubin with free thyroxine is modified by insulin resistance in euthyroid subjects. PLoS One  2014; 9: e90886 Google Scholar CrossRef Search ADS PubMed  31 Hallan S, Astor B, Lydersen S. Estimating glomerular filtration rate in the general population: the second Health Survey of Nord-Trondelag (HUNT II). Nephrol Dial Transplant  2006; 21: 1525– 1533 Google Scholar CrossRef Search ADS PubMed  32 Levey AS, Bosch JP, Lewis JB et al.   A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med  1999; 130: 461– 470 Google Scholar CrossRef Search ADS PubMed  33 Levey AS, Stevens LA, Schmid CH et al.   A new equation to estimate glomerular filtration rate. Ann Intern Med  2009; 150: 604– 612 Google Scholar CrossRef Search ADS PubMed  34 Ferguson MA, Waikar SS. Established and emerging markers of kidney function. Clin Chem  2012; 58: 680– 689 Google Scholar CrossRef Search ADS PubMed  35 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials  1986; 7: 177– 188 Google Scholar CrossRef Search ADS PubMed  36 Asvold BO, Bjoro T, Vatten LJ. Association of thyroid function with estimated glomerular filtration rate in a population-based study. The HUNT Study. Eur J Endocrinol  2011; 164: 101– 105 Google Scholar CrossRef Search ADS PubMed  37 Rhee CM, Kalantar-Zadeh K, Streja E et al.   The relationship between thyroid function and estimated glomerular filtration rate in patients with chronic kidney disease. Nephrol Dial Transplant  2015; 30: 282– 287 Google Scholar CrossRef Search ADS PubMed  38 Mandelli S, Riva E, Tettamanti M et al.   Mortality prediction in the oldest old with five different equations to estimate glomerular filtration rate: the health and anemia population-based study. PLoS One  2015; 10: e0136039 Google Scholar CrossRef Search ADS PubMed  39 Kaptein EM. Thyroid hormone metabolism and thyroid diseases in chronic renal failure. Endocr Rev  1996; 17: 45– 63 Google Scholar CrossRef Search ADS PubMed  40 Warner MH, Beckett GJ. Mechanisms behind the non-thyroidal illness syndrome: an update. J Endocrinol  2010; 205: 1– 13 Google Scholar CrossRef Search ADS PubMed  41 Wang X, Lewis J, Appel L et al.   Validation of creatinine-based estimates of GFR when evaluating risk factors in longitudinal studies of kidney disease. J Am Soc Nephrol  2006; 17: 2900– 2909 Google Scholar CrossRef Search ADS PubMed  42 Hallan SI, Ritz E, Lydersen S et al.   Combining GFR and albuminuria to classify CKD improves prediction of ESRD. J Am Soc Nephrol  2009; 20: 1069– 1077 Google Scholar CrossRef Search ADS PubMed  43 Verhave JC, Gansevoort RT, Hillege HL et al.   An elevated urinary albumin excretion predicts de novo development of renal function impairment in the general population. Kidney Int Suppl  2004; 92: S18– S21 Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Nephrology Dialysis TransplantationOxford University Press

Published: Apr 18, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off