Renal “hyperfiltrators” are at elevated risk of death and chronic diseases

The baseline study sample is formed by consecutive outpatients and patients admitted in the first quarter of year 2010 to the S. Ersek Center for Cardiovascular Surgery who had no missing values for serum creatinine concentrations. This sample was drawn from a total of 1900 patients applying in the stated period. For reasons of lower risk of cardiopulmonary events or confounding problems of potential comorbidity, subjects aged <40, or >72 years and those with greater than mild renal dysfunction (<60 ml/min per 1.73 m²) (n = 207), decompensated heart failure (n = 120) and uncontrolled diabetes (HbA1c >10%; n = 92) were excluded, as were subjects with missing creatinine values (n = 494). This left 789 individuals to form the study sample. Follow-up consisted of up to 3.0 years. The study was approved by the Institutional Ethics Committee of the Dr. Siyami Ersek Training Research Hospital Center. Data recorded in the hospital charts were obtained.

As this study is a retrospective study, informed consent was not obtained from the patients.

Blood pressure (BP) was measured with an aneroid sphygmomanometer (Erka, Germany) in the sitting position on the right arm, and the mean of at least two recordings 5 min apart was registered. Body mass index (BMI) was computed from values of weight/height squared. Never, former and current smokers formed the categories of cigarette smoking.

Sixty-four percent of the study participants underwent two-dimensional and Doppler echocardiography by an experienced specialist using a Vivid 3 (GE Healthcare Systems, Piscataway, New Jersey, USA) device. Left ventricular ejection fraction (LVEF) was calculated using the modified Simpson’s method.

GFR (in ml/min per 1.73 m²) was estimated using a) the equation proposed by CKD-EPI for white females and males [14] which uses truncation for low creatinine values, and b) the abbreviated equation developed by the MDRD [13] = (186.3 * (serum creatinine)^–1.154 * age^–0.203 * 0.742 [for women]). Hypertension was defined as a blood pressure ≥140 mmHg and/or ≥90 mmHg, and/or use of antihypertensive medication. Diabetes type-2 was diagnosed with the criteria of the American Diabetes Association [15]. CHD was diagnosed by a history of myocardial infarction and revascularization, by angiographic greater than 50% stenosis of at least one major coronary artery.

Information on cause of death was drawn from hospital records during follow-up, from records of the nation-wide Identity Sharing System, or from first-degree relatives. Follow-up for mortality was complete via the Identity Sharing System which registers all deaths by using identity numbers for each citizen; causes were identified from hospital records or by phone from first-degree relatives. Information on nonfatal events at follow-up was obtained (in 70%) from hospital records at their return visits to the medical center, and the remainder by phone from close relatives of patients, Stroke was defined as affected impairment of brain function having resulted in an inability to move a limb, or inability to understand or formulate speech. Congestive heart failure (HF) was diagnosed when typical symptoms (e.g. breathlessness, ankle swelling, and fatigue) and signs (e.g. elevated jugular venous pressure, pulmonary crackles, and displaced apex beat) due to an abnormality of cardiac structure. Chronic obstructive pulmonary disease (COPD) designated patients who had no evidence of this at baseline but received this diagnosis in the follow-up period in our out-patient chest clinic, based on obstructive findings in pulmonary function testing and in whom bronchodilator drugs were initiated.

Descriptive parameters were shown as mean (±standard deviation [SD]) or in percentages. Due to skewed distribution, geometric means were used for CRP values. Two-sided t-tests and Pearson’s chi-square tests were used to analyze the differences between means and proportions of variables within quartiles of eGFR and by pairwise comparisons with post hoc Tukey homogeneous subsets (HSD) tests. The development of multiple cardiopulmonary events in an individual was counted only as one subject, or in the case of death as death alone. Rather than defining fixed cutoffs for hyperfiltration for which unanimity is lacking, we formed sex-specific quartiles of eGFR defined by each equation. Cutoff points were 60–88, up to 99, up to 110, and >110 ml/min per 1.73 m² for CKD-EPI, while these were <83.6/<82.6, up to 100/105, up to 132/141.9, >132/≥142 ml/min per 1.73 m², respectively, for men/women in the MDRD formula. A multiple linear regression model, selected from findings of univariate correlation, was analyzed to seek best independent covariates of eGFR. Logistic regression analyses were used for combined endpoint of death and incident cardiovascular and pulmonary events (MI, HF, stroke and COPD) for eGFR quartiles formed by the two sets of cutoffs. Risk estimates (RR) and 95% confidence intervals (CI) were obtained in models that adjusted for sex, age and relevant confounders, expressed in terms of 1-SD increment. A value of p < 0.05 on the two-tail test was considered statistically significant. Statistical analyses were performed using SPSS-10 for Windows.

Five variables showing highest univariate correlations with eGFR were included in a multiple linear regression model to determine its independent covariates (Table 3). Using eGFR derived from both equations in highly significant regression models, serum uric acid levels proved a highly significant inverse covariate in the “healthy” group, independent of BMI and apart from age. LVEF was associated inversely with eGFR, significantly in the “healthy” group, tended weakly so in the CHD group.

Table 4 shows predictors of combined outcome (death and nonfatal cardiopulmonary events) among the groups without and with CHD at baseline using the same regression model with eGFR derived from two equations. Among the “healthy” group, beyond age and former smoking, highest eGFR quartile with the MDRD equation significantly predicted the combined outcome at a RR of 4.00 (95%CI 1.34; 11.9) compared to the normal Q2, while sex, BMI and presence of hypertension did not reach significance. RR with the CKD-EPI was 2.94 (whereby p-trend =0.048), or attained a p = 0.006 when expressed using Q1 as referent (Table 4). Highest eGFR quartile significantly predicted outcome among patients with CHD at an RR 2.72 (95% CI 1.08; 6.81) yet only with the MDRD estimate, not with that of CKD-EPI.

Individuals included in CKDI-EPI-derived eGFR Q2 or Q3 may be considered to represent normal renal function, displaying a range of 88–110 ml/min per 1.73 m². Healthy females had eGFR similar to men while females with CHD had even somewhat higher eGFR than men, although in the general TARF sample [16] they had one-tenth lower values than men. This current finding is consistent with Turkish women being more susceptible to autoimmune activation than men, with resultant apparent “hyperfiltration” [12].

The current study demonstrates that, in a sample of subjects undergoing health checks, men as well are subject to Lp (a)-activated autoimmunity comprising serum creatinine. By showing similar distribution of the two GFR estimates across the quartiles in men with CHD, this study supported the conclusion by the TARF [9, 10] that, in contrast to females, serum creatinine is little involved in an autoimmune complex in men during the development of CHD, but is rather so in healthy men mainly regarding other outcomes such as heart failure, stroke and COPD.

Although tobacco smoking is a major risk factor for COPD, recent evidence suggests that other risk factors are important as well, especially in developing countries [17]. One-third of patients with COPD are estimated to be never smokers but are thought to be exposed to biomass fumes [18‐20], air pollutants, etc. Beyond this, enhanced pro-inflammatory state associated with obesity might result also in COPD among non-smokers, a tendency reported among Korean men [20]. On this postulate, we included COPD in this study as a potentially informative outcome and observed that all 5 cases of newly developing COPD in the “healthy” group belonged to baseline eGFR quartiles 3 and 4.

It is worth pointing out in this context that discontinuance of smoking was a significant predictor of combined outcome events in the multivariable regression model, suggesting inhibition of autoimmune processes by active smoking and being in line with the findings and interpretation regarding autoimmune type-2 diabetes developing in Norwegian men and women [21].

That low serum creatinine was not due to low muscle mass associated with end-stage renal disease, or advanced heart failure, can be inferred from the younger age, similar BMI and LVEF of hyperfiltrators compared to participants in the remaining categories. Furthermore, a “reverse causality” is highly unlikely due to the prospective design of the study. Multiple linear analyses indicated that BMI was not independently related to hyperfiltration. Conversely, these patients were not undernourished, as evidenced by their being overweight in general. They were associated with atherogenic (high triglyceride/low HDL-cholesterol) dyslipidemia and exhibited, notably, lower concentrations of LDL- and total cholesterol than subjects with normal eGFR, similar to the large Swedish AMORIS study cohort [22] in which the overwhelming majority with a normal eGFR had 12% lower LDL- and total cholesterol values than those with mildly reduced eGFR. Yet when classified into quartiles, multivariable adjustment revealed in the highest eGFR quartile higher risk for all-cause mortality than the optimal risk quartile (>96 ml/min vs. 85–96 ml/min per 1.73 m²) and for incident MI than all the remaining eGFR quartiles.

Lower uric acid values among “hyperfiltrators” suggest the concomitant involvement in the immune complex of uric acid as well, for which some evidence was previously found in Turkish women [23, 24]. This consideration is in agreement with a large prospective study in Taiwan reporting that sex- and age-adjusted hazard ratios of uric acid levels for overall mortality were increased, also for uricemia lower than 5 mg/dl [25]. Our finding that “hyperfiltrators” were both younger and had lower serum uric acid are in line with recently reported observations among Turks that uric acid was independently and linearly associated with serum total phospholipids in men and women without MetS and that the concentration of phospholipids on HDL predicted the development of MetS [24], thus supporting a notion that –similar to creatinine– reduced uric acid level may be due to partial inability to assay, secondary to autoimmune activation. Puddu and Menotti [26] stressed that “serum uric acid might be considered for risk predictive purposes to be incorporated into software instruments aimed at indexing and/or following up individuals for primary preventive purposes.”

“Hyperfiltrators”, characterized by high serum triglycerides, suggested the presence of low circulating Lp(a) –by virtue of having damaged epitope that partly escapes immunoassay. Such an autoimmune “pattern” has been reported in the TARF study [12, 27]. Reduced levels of LDL-cholesterol have been reported to similarly precede the clinical onset of rheumatoid arthritis [28]. Epidemiological data showing paradoxical inverse association of lipid levels with cardiovascular disease risk in rheumatoid arthritis patients have been reviewed [29]. Furthermore, high serum triglycerides headed the determinants of a new cardiovascular event n 700 Swedish early rheumatoid arthritis patients [30].

Noteworthy is that the characteristics of atherogenic dyslipidemia and reduced LVEF of “hyperfiltrators” became diluted in the subset with baseline CHD, since CHD patients with normal renal function not only had diminished LVEF but also low total and LDL-cholesterol and a higher proportion of atherogenic dyslipidemia, eliminating the differences from the “hyperfiltrators”.

Present findings support the hypothesized notion [12, 27] that autoimmune activation associated with enhanced low-grade inflammation is a common mechanistic pathway leading to diverse chronic diseases, foremost to diabetes, CHD and CKD. Serum creatinine, a tripeptide, may undergo oxidative damage to its epitope alike circulating a), be perceived as a foreign body, partly lose assayability and may aggregate to autoimmune components such as apoA-I [11] and/or adiponectin. Immune complexes with ß₂-glycoprotein I-Lp (a) have been reported in patients with coronary artery disease [31]. The related individual is necessarily categorized into the lowest creatinine (or highest eGFR quantile), despite likely harboring a milieu of oxidative stress, HDL dysfunction and autoimmune activation associated with elevated cardiorenal-metabolic risk.

The reduced LVEF, along with our observation of HF as the most common consequence among “healthy” “hyperfiltrators”, suggests that such autoimmune activation frequently leads to HF the prevalence of which has been increasing in the community [32]. Cancer formed the second frequent death cause, being in line with a recent community-based case–control study that patients with heart failure have an increased risk of incident cancer [33], suggesting a common underlying mechanism.

The magnitude of relative risk it conferred in this sample, and the substantial proportion in the population at large, necessitate the identification of potential “hyperfiltrators” in the community which has both methodological and clinical implications. One implication concerns the equation selected for estimating GFR, since the CKD-EPI [14], more suited to estimate creatinine-based GFR for the general population, tends to conceal “hyperfiltrators” by truncating creatinine values <0.7/<0.9 mg/dl. The MDRD estimate has the advantage of disclosing apparent “hyperfiltrators”, as can be inferred from analyzing eGFR quartiles in the current as well as the AMORIS study [22], commented above.

For clinically suspecting “hyperfiltrating” individuals, we may propose as a rough guide for 55-year old Turks an eGFR of >120/110 ml/min per 1.73 m² (or serum creatinine <0.72 and <0.60 mg/dl) in men and women, respectively. The need for better understanding the mechanisms involved in renal hyperfiltration has been emphasized in order to develop new therapeutic strategies [1]. Much further research is warranted in this area in different ethnicities.

The size of the study sample and the duration of the follow-up period are relatively limited which, however, did not preclude the emergence of significant results in the sample. Applicability of current findings, especially the extent of such in other ethnicities remains to be confirmed, but arguments may apply to similar findings among Swedes [22]. Given that underlying pathways examined here had already been published in representative samples of Turkish adults, we generated congruent results in patients suspected of heart disease undergoing health checks. Findings regarding COPD, obviously, may be considered only as preliminary and hypothesis generating. Strength of the study is its unique longitudinal design, as well as the stratified analyses of a sample susceptible to pro-inflammatory state, the availability of various clinical, lipid and inflammation biomarkers, including LVEF, thus supporting and confirming a novel hypothesis. The salient information provided by the concomitant evaluation of GFR estimates and their relative predictive value of selected outcomes is thus far unique.