Association of Blood Lead Levels With Mortality in Patients on Maintenance Hemodialysis

Transcription

1 CLINICAL RESEARCH STUDY Association of Blood Lead Levels With Mortality in Patients on Maintenance Hemodialysis Ja-Liang Lin, MD, a Dan-Tzu Lin-Tan, RN, a Ching-Wei Hsu, MD, a Tzung-Hai Yen, MD, a Kuan-Hsing Chen, MD, a Hsiang-Hao Hsu, MD, a Tai-Chin Ho, MT, a Kuang-Hong Hsu, PhD b a Department of Nephrology and Division of Clinical Toxicology, Chang Gung Memorial Hospital, Taipei, Lin-Kou Medical Center, Taoyuan, Chang Gung University and School of Medicine, Taoyuan, Taiwan, ROC; b Laboratory of Epidemiology, Department of Health Care Management, Chang Gung University, Taoyuan, Taiwan, ROC. ABSTRACT BACKGROUND: The association between blood lead s and mortality in patients on maintenance hemodialysis remains unclear. METHODS: A cross-sectional and 18-month prospective study included 927 patients on maintenance hemodialysis. Baseline variables and blood lead s were measured before hemodialysis and categorized as 3 equal groups: high ( g/dl), middle ( g/dl), and low ( 8.51 g/dl). Mortality and cause of death were recoded for longitudinal analyses. RESULTS: At baseline, after related variables were adjusted, logarithmic transformation of blood lead was negatively related to log ferritin and positively related to the vintage of hemodialysis and the percentage of urban area patients. By the end of the follow-up, 59 patients had died. Kaplan Meier survival analysis showed that the high blood lead group had greater mortality than the low blood lead group (log-rank test, P.001). After adjustment for potential variables, Cox multivariate analysis demonstrated that by using the low blood lead as the reference, high blood lead s were associated with increased hazard ratios (HRs) for all-cause (HR 4.70; 95% confidence interval [CI], ; P.003), cardiovascular-cause (HR 9.71; 95% CI, ; P.005), and infection-cause (HR 5.35; 95% CI, ; P.046) 18-month mortality in patients on maintenance hemodialysis. Moreover, there was a significant trend (P.032) of HRs for all-cause mortality among the 3 study groups. CONCLUSION: High blood lead is associated with increased HRs for all-cause, cardiovascular-cause, and infection-cause 18-month mortality in patients on maintenance hemodialysis Elsevier Inc. All rights reserved. The American Journal of Medicine (2011) 124, KEYWORDS: Blood lead s; Maintenance hemodialysis; Mortality Funding: National Science Council, Republic of China. Contract No. NSC B-182A-042-MY3. Conflict of Interest: None of the authors have any conflicts of interest associated with the work presented in this manuscript. Authorship: All authors had access to the data and played a role in writing this manuscript. Requests for reprints should be addressed to Ja-Liang Lin, MD, Department of Nephrology and Division of Clinical Toxicology, Chang Gung Memorial Hospital, 199, Tung-Hwa North Road, Taipei, Taiwan, ROC. address: jllin99@hotmail.com. Epidemiologic studies have suggested that blood lead is associated with all-cause, cardiovascular-cause, and cancer-cause mortality 1,2 and may increase blood pressure 3 in the general population. Moreover, blood lead has been demonstrated to accelerate age-related renal function impairment in the general population. 4,5 Recent investigations also have shown that environmental exposure to lead is associated with progressive renal insufficiency 6-8 and that chelation therapy may retard the progression of renal insufficiency 8-10 in patients with chronic kidney disease. Previous studies report that lead accumulates in patients with end-stage renal disease, and that blood lead is related to hemoglobin, blood pressure, and intact parathyroid hormone s in patients on maintenance hemodialysis. 11,12,13 Because no adjustment for related variables and small sample size are noted in these studies, these results are inconclusive. Two studies suggested that blood lead was associated with increased risks for mortality in a small group of diabetic patients on maintenance hemodialysis 14 and patients on chronic peritoneal dialysis. 15 However, the /$ -see front matter 2011 Elsevier Inc. All rights reserved. doi: /j.amjmed

2 Lin et al Lead and Mortality in Patients on MHD 351 clinical significance of blood lead in patients on maintenance hemodialysis remains unclear. Moreover, in Taiwan 16 and the United States, 17 approximately 50% of all the reported deaths of patients on maintenance hemodialysis are due to cardiovascular complications. It remains unknown whether blood lead relates to cardiovascular-cause mortality in patients on maintenance hemodialysis. In this multicenter study, the relationship between baseline blood lead and variables was evaluated and 18-month follow-up data were analyzed to determine whether blood lead increases mortality risk in patients on maintenance hemodialysis. MATERIALS AND METHODS This clinical study was performed in accordance with the Declaration of Helsinki and approved by the medical ethics committee of Chang Gung Memorial Hospital, Taipei. Informed consent was obtained from all patients. CLINICAL SIGNIFICANCE Patients All study patients were recruited from 3 hemodialysis centers of Chang Gung Memorial Hospital in Taipei, Lin-Kou, and Toayuan. Only patients on maintenance hemodialysis who were aged 18 years or more and who received hemodialysis for 6 months were enrolled in the study. Patients with malignancies and obvious infectious diseases and those who had been hospitalized or had undergone surgery within 3 months before the investigation were excluded. In addition, a questionnaire was arranged to survey patients who had a history of occupational exposure to lead or previous lead intoxication or those living in lead-contaminated areas. These patients also were excluded. Patients with diabetes mellitus were defined as those with a previous diagnosis of diabetes mellitus or a fasting blood glucose 126 mg/dl twice subsequently. Most of the patients underwent 4 hours of hemodialysis, 3 times per week. Hemodialysis for these patients was carried out using single-use, hollowfiber dialyzers equipped with modified cellulose, polyamide, or polysulfone membranes. The dialysate used in all cases had a standard ionic composition with bicarbonatebased buffer. We noted the incidence of cardiovascular diseases, including cardiovascular disease (International Classification of Diseases 10th Revision [ICD-10] codes I00-I99), myocardial infarction (ICD-10 codes I20-I25), and stroke (ICD-10 codes I60-I69). Patients with hypertension were defined as those with at least 2 blood pressure measurements 140/90 mm Hg or taking antihypertensive drugs regularly to control their blood pressure. In patients on maintenance hemodialysis, after adjustment for related variables, a high blood lead ( g/dl) is associated with increased hazard ratios (9.7-fold for cardiovascular-cause, 5.4-fold for infection-cause, and 4.7-fold for all-cause mortality) compared with a low blood lead ( 8.51 g/dl). Study results lend support to reduce environmental exposure to lead, avoid iron-deficiency status, and explore a possible chelation therapy for these patients. Measurement of Blood Lead Levels To exclude the possibility that patients on maintenance hemodialysis were exposed to lead through the contamination of water or dialysate during hemodialysis, we collected at least 2 samples of water and dialysate from outlets of reverse osmosis systems and inlets of dialysate of dialyzers in leadfree plastic bottles from each hemodialysis center. Blood lead s were measured using a method described previously. 18 Lead s were measured using an electrothermal atomic-absorption spectrometer (SpectrAA-200Z; Varian, Lexington, Mass) with Zeeman background correction and a L vov platform. A certified commercially prepared product (Seronorm Trace Elements; Sero AS, Billingstads, Norway) was used to determine intra-batch accuracy and confirm the inter-batch standardization. The coefficient of variation for lead measurement was 5.0%. External quality control was maintained by participating in the National Quality Control Program conducted by the government. Blood lead s of each patient were measured 2 times with a 3-month interval. The mean blood lead s were stratified into 3 equal groups: low blood lead ( 8.51 g/dl), middle blood lead ( g/dl), and high blood lead ( g/dl). Laboratory Data Blood samples were collected from the arterial end of the vascular access immediately after initiating a 2-day hemodialysis at mid-week, centrifuged, and stored at 80 C for further analysis and measurement of blood lead s and biochemical data. Serum albumin, creatinine, transferrin saturation, and normalized protein catabolic rate were assayed and recorded as nutritional markers. High-sensitivity C-reactive protein was a marker of inflammation. Serum high-sensitivity C-reactive protein concentrations were measured by immunonephelometry (Nanopia CRP; Daiichi, Inc, Tokyo, Japan). All other markers were measured in standard laboratory procedures using an automatic analyzer. The normalized protein catabolic rate was calculated using validated equations and normalized to actual body weight. 19 The clearance of urea through dialysis was expressed as Kt/V, using the Daugirdas 20 method. Serum calcium s were corrected as follows: corrected calcium s (milligrams/deciliter) serum calcium s serum albumin s (milligrams/deciliter). Abnormal high-sensitivity C-reactive protein was defined as 3.0 mg/l, which corresponded to the

3 352 The American Journal of Medicine, Vol 124, No 4, April 2011 in patients on maintenance hemodialysis with an increased cardiovascular risk. 21,22 Abnormal albumin was defined as 3.6 g/dl, which is the low limit of normal serum albumin in our hospital and the 10th percentile of the values obtained for Americans in the Third National Health and Nutrition Examination survey of Americans. 22,23 Follow-up for 18 Months Every death during the follow-up period was reviewed and assigned an underlying cause by physicians not involved in this study. Cardiovascular-cause death is defined as cardiovascular disease (ICD-10 codes I00-I99), myocardial infarction (ICD-10 codes I20-I25), and stroke (ICD-10 codes I60-I69). In patients who died in the hospital, cardiovascular or infection events during follow-up were obtained from discharge diagnosis and death certificates in charters in the hospital. In the case of an out-of-hospital death, family members were interviewed by telephone to ascertain the circumstances surrounding the death. Statistical Analysis Unless otherwise stated, continuous variables are expressed as means standard deviation, and categoric variables are expressed as numbers or percentages for each item. Comparisons among the 3 tertile groups of patients were analyzed by using a trend test. A logarithmic conversion was performed for non-normal distribution of variables, such as blood lead (log blood lead ), intact parathyroid hormone (log intact parathyroid hormone), ferritin (log ferritin), and high-sensitivity C-reactive protein (log high-sensitivity C-reactive protein) s. To elucidate the association between baseline blood lead and variables, we performed both simple linear regression analysis and stepwise multiple linear regression analysis for all potential related variables. Kaplan Meier curves were compared by using the log-rank test. Cox proportional hazard models were performed to calculate the relative risk of death by obtaining hazard ratios (HRs) at 95% confidence intervals (CIs). A univariate Cox model was used to assess all previously identified variables for mortality of in the literature, and variables with P.05 were entered into the final multivariate Cox models. Because hemodialysis vintage was significantly different among the 3 study groups, we entered the hemodialysis vintage in each multivariate Cox model. Because most lead exists in red blood cells, blood lead is corrected by hemoglobin (corrected blood lead ): in men, corrected blood lead blood lead 14.0/hemoglobin s; in women, corrected blood lead blood lead 12/ hemoglobin s. We assessed the corrected blood lead in the multivariate Cox models. Moreover, we also performed the multivariate Cox models to assess the relation between blood lead and mortality in nondiabetic patients on maintenance hemodialysis. The multivariate Cox analysis was also calculated to assess the relation between blood lead and mortality for all patients, including 19 patients with previous exposure to lead. To clarify the possible confounding effect of living in an urban area, we also categorized patients as living in an urban area or a countryside area. Multivariate Cox analysis was performed to assess whether high blood lead was associated with an increased risk for mortality. P.05 was significant for all statistical tests. Data were analyzed using SPSS software version 12.0 for Windows 95 (SPSS, Inc, Chicago, Ill). RESULTS Patient Characteristics A total of 927 patients (470 male and 457 female) receiving maintenance hemodialysis with a mean hemodialysis duration of years were enrolled in the study. Table 1 lists the clinical characteristics of these patients, including age, gender, body mass index, and biological and biochemical data. The mean patient age was years (range years). The median blood lead was 10.4 g/dl (range g/dl). Table 1 also lists patient characteristics for the tertile of 3 blood lead groups. Patients in the high blood lead group (n 309) (median blood lead 16.4 g/dl, range g/dl) had a trend of a longer vintage of hemodialysis and showed higher blood flow and dialysate flow rates, higher Kt/V (Daugirdas) urea values, and intact parathyroid hormone s than the other 2 groups. A higher percentage of the patients in the high blood lead group also lived in an urban area, had diabetes, used biocompatible membrane of dialyzers, and had a residual urine amount 100 ml/d, compared with the low (n 309) (median 6.3 g/dl, range g/dl) and middle (n 309) (median 10.4 g/dl, range g/dl) blood lead groups. Moreover, patients in the high blood lead group had a trend of lower serum albumin and ferritin s than the other groups. The groups did not differ in other baseline variables (Table 1). Lead Levels in Water and Dialysate and Blood Lead Levels of Patients on Maintenance Hemodialysis Lead s in all the water and dialysate samples (n 12) were less than 2 g/l, which was less than the American Association for Advancement of Medical Instrumentation standard ( 50 g/l). Correlations Between Blood Lead Levels or Corrected Blood Lead Levels and Baseline Data in Patients on Maintenance Hemodialysis After adjustment for related variables, stepwise multiple linear regression analysis showed that the percentages of those living in an urban area, the vintage of hemodialysis, and the log ferritin were associated with log blood lead in patients on maintenance hemodialysis (Table 2). Similarly, gender, living in an urban area, vintage of hemodialysis, hemoglobin, and serum albumin were related to log-corrected blood lead s in these patients (Table 2).

4 Lin et al Lead and Mortality in Patients on MHD 353 Table 1 Baseline Characteristics of Patients on Maintenance Hemodialysis by Tertile of Blood Lead Levels (N 927) Characteristics Low Blood Lead Level Group ( 8.51 g/dl) (n 309) Middle Blood Lead Level Group ( g/ dl) (n 309) High Blood Lead Level Group ( g/dl) (n 309) P Demographics Age (y) Female sex 152 (49.2) 154 (49.2) 151 (48.9).936 Body mass index (kg/m 2 ) Smoking 56 (18.1) 55 (17.8) 57 (18.4).917 Urban area 9 (2.9) 25 (8.1) 141 (45.6).001 Education s (high school or 113 (36.6) 111 (35.9) 123 (39.8).524 more) Previous cardiovascular diseases 12 (3.9) 18 (5.8) 16 (5.2).439 Diabetes mellitus 83 (26.9) 74 (23.9) 43 (13.9).001 Hypertension 125 (40.5) 122 (39.5) 117 (37.9).510 Dialysis-related data Hemodialysis duration (y) Use of fistula 246 (79.6) 258 (83.5) 250 (80.9).686 Use of BCM dialyzers 220 (71.2) 226 (73.1) 244 (79.0).025 Blood flow rate (ml/min) Dialysate flow rate (ml/min) Erythropoietin (U/kg/wk) Kt/V (Daugirdas) npcr (g/kg/d) Residual daily urine 100 ml 223 (72.2) 255 (82.5) 253 (81.9).004 Biochemical data Hemoglobin (g/dl) Albumin(g/dL) Albumin 3.5 g/dl 15 (4.9) 26 (8.4) 27 (8.7).055 Creatinine (mg/dl) Transferrin Saturation (%) Ferritin ( g/l) 374 (7-3571) 300 (9-2618) 285 (8-2688).001 C-Ca (mg/dl) Phosphate (mg/dl) Intact parathyroid hormone (pg/ml) 115 (0-1655) 121 (0-1727) 180 (1-1754).001 Cardiovascular risks Cholesterol (mg/dl) HDL-C (mg/dl) LDL-C (mg/dl) Triglyceride (mg/dl) HsCRP (mg/l) 2.82 ( ) 2.89 ( ) 2.80 ( ).972 HsCRP 3.0 mg/l 148 (47.9) 147 (47.6) 144 (46.6).721 Cardiothoracic ratio (%) BCM biocompatible membrane; C-Ca corrected calcium; HD hemodialysis; HDL-C high-density lipoprotein cholesterol; HsCRP high-sensitivity C-reactive protein; Log IPTH logarithmic transformation of intact parathyroid hormone; npcr normalized protein catabolic rate; LDL-C lowdensity lipoprotein cholesterol. P.05 showed significant trends among groups. Non-normal distribution data were presented as median (minimum, maximum). Hypertension: blood pressure 140/90 mm Hg twice at least or regularly taking antihypertension drugs. Previous cardiovascular diseases included cerebrovascular disease, coronary arterial disease, congestive heart failure, or peripheral vascular disease in the past. Diabetes mellitus was defined as a previous diagnosis of diabetes mellitus or fasting glucose s 126 mg/dl twice subsequently. Kaplan Meier and Multivariate Cox Regression Analysis for 18-Month Mortality in Patients on Maintenance Hemodialysis At the end of the 18-month observation period, the results showed the following: Twenty-nine patients (29/59; 49.2%) died of cardiovascular diseases, 27 patients (27/59, 45.8%) died of infection, 2 patients (2/59, 3.4%) died of malignancies, and 1 patient (1/59, 1.7%) died of unknown cause. Of these 59 patients (59/927, 6.4%), 31 (31/309, 10.0%) with a high blood lead, 20 (20/309, 6.5%) with a middle blood lead, and 8 (8/309, 2.6%) with a low blood lead died. A total of 803 patients completed the 18-month

5 354 The American Journal of Medicine, Vol 124, No 4, April 2011 Table 2 Determinants of Blood Lead Levels or Corrected Blood Lead Levels in Patients on Maintenance Hemodialysis (N 927) Stepwise Multiple Regression Analysis Variables Beta coefficient SE P Log blood lead s Urban area (yes 1) Hemodialysis vintage (y) Log ferritin (g/dl) Log-corrected blood lead s Gender (male 1) Urban area (yes 1) Hemodialysis vintage (y) Hemoglobin (g/dl) Serum albumin (g/dl) Log logarithmic transformation; SE standard error. follow-up (Figure 1). Kaplan Meier survival analysis showed that the high blood lead group had greater mortality than the low blood lead group (log-rank test, P.001; Figure 2). After adjusting for baseline variables with P.05 in the univariate Cox proportional hazards analysis and hemodialysis vintage, with the low blood lead as the reference, high blood lead (HR 4.70; 95% CI, ; P.003) was associated with increased HRs for all-cause 18-month mortality in patients on maintenance hemodialysis (Table 3). Similarly, middle blood lead (HR 3.70; 95% CI, ; P.044) and high blood lead (HR 9.71; 95% CI, ; P.005) were associated with increased HRs for cardiovascular-cause mortality in these patients (Table 4). Moreover, high blood lead (HR 5.35; 95% CI, ; P.046) also was associated with increased HRs for infection-cause 18-month mortality in these patients (Table 5). There was a significant trend (P.032) of HRs for all-cause mortality among the 3 study groups. Figure 1 Flow chart shows enrollment and status of patients. HD hemodialysis.

6 Lin et al Lead and Mortality in Patients on MHD 355 Figure 2 Kaplan Meier survival analysis shows that all-cause mortality rate (10.0%, 31/309) in the high blood lead group is higher than in the low blood lead group (2.6%, 8/309) after 18-month prospective follow-up (log-rank tests, ; P.001). BLL blood lead. After adjusting for hemodialysis vintage and potential variables that were significant (P.05) in univariate Cox analysis, with the low corrected blood lead ( g/dl) as the reference group, high corrected blood lead Table 3 Cox Multivariate Regression Analysis of Hazard Ratios for All-Cause 18-Month Mortality in All Patients on Maintenance Hemodialysis, According to Baseline Tertile of Blood Lead Level or Corrected Blood Lead Level and Variables That Had P.05 in Univariate Cox Analysis Variables* Blood lead Low blood lead as the reference ( 1) Middle blood lead Multivariate Hazard Analysis HRs (95% CIs) P ( ).094 High blood lead 4.70 ( ).003 Corrected blood lead Low corrected blood lead as the reference ( 1) Middle corrected blood lead High corrected blood lead ( ) ( ).006 CI confidence interval; HR hazard ratio. *After adjustment for age, diabetes mellitus, education (high school or more), hemodialysis vintage, using biocompatible membrane of dialyzers, normalized protein catabolic rate, daily residual urine 100 ml, hemoglobin, serum albumin, high-density lipoprotein cholesterol, creatinine, phosphate, cardiothoracic ratio, logarithmic transformation of ferritin, intact parathyroid hormone, and high-sensitivity C-reactive protein. Diabetes mellitus was defined as a previous diagnosis of diabetes mellitus or fasting glucose s 126 mg/dl twice subsequently. ( g/dl) was associated with increased HRs for all-cause (HR 4.98; 95% CI, ; P.006) (Table 3), cardiovascular-cause (HR 7.35; 95% CI, ; P.020) (Table 4), and infection-cause (HR 4.72; 95% CI, ; P.047) (Table 5) 18-month mortality in patients on maintenance hemodialysis. Table 4 Cox Multivariate Regression Analysis of Hazard Ratios for Cardiovascular-Cause 18-Month Mortality in all Patients on Maintenance Hemodialysis, According to Baseline Tertile of Blood Lead Level or Corrected Blood Lead Level and Variables That Had P.05 in Univariate Cox Analysis Variables* Blood lead Low blood lead as the reference ( 1) Middle blood lead Multivariate Hazard Analysis HRs (95% CIs) P ( ).044 High blood lead 9.71 ( ).005 Corrected blood lead Low corrected blood lead as the reference ( 1) Middle corrected blood lead High corrected blood lead ( ) ( ).020 CI confidence interval; HR hazard ratio. *After adjustment for age, previous cardiovascular diseases, education (high school or more), hemodialysis vintage, using fistula, normalized protein catabolic rate, hemoglobin, serum albumin, creatinine, cardiothoracic ratio, and logarithmic transformation of high-sensitivity C-reactive protein. Previous cardiovascular diseases included stroke, myocardial infarction, peripheral vascular disease, and congestive heart failure.

7 356 The American Journal of Medicine, Vol 124, No 4, April 2011 Table 5 Cox Multivariate Regression Analysis of Hazard Ratios for Infection-Cause 36-Month Mortality in All Patients on Maintenance Hemodialysis, According to Baseline Tertile of Blood Lead Level or Corrected Blood Lead Level and Variables That Had P.05 in Univariate Cox Analysis Variables* Blood lead Low blood lead as the reference ( 1) Middle blood lead Multivariate Hazard Analysis HRs (95% CIs) Moreover, if patients with diabetes were excluded, only nondiabetic patients on maintenance hemodialysis were assessed. After adjusting for hemodialysis vintage and potential variables, with the low blood lead as the reference, high blood lead was associated with increased HRs for all-cause (HR 5.85; 95% CI, ; P.027) 18- month mortality in nondiabetic patients on maintenance hemodialysis. If 19 patients with previous exposure to lead were included in this study (n 946), the multivariate Cox analysis showed, after adjustment for potential variables and hemodialysis vintage, with low blood lead ( 8.56 g/dl) as the reference, that high blood lead ( g/dl) was associated with increased HRs for all-cause (HR 4.76; 95% CI, ; P.003) 18-month mortality in all patients on maintenance hemodialysis. Forty-four deaths occurred in the countryside area during the 18-month follow-up. The multivariate Cox analysis showed, after adjustment for potential variables and hemodialysis vintage, with the low blood lead ( 7.72 g/ dl) as the reference, that high blood lead ( g/dl) was associated with increased HRs for all-cause (HR 2.94; 95% CI, ; P.048) mortality in all countryside patients (n 752). There was no association between blood lead and mortality in urban area patients. P ( ).662 High blood lead 5.35 ( ).046 Corrected blood lead Low corrected blood lead as the reference ( 1) Middle corrected blood lead High corrected blood lead ( ) ( ).047 CI confidence interval; HR hazard ratio. *After adjustment for age, diabetes mellitus, hemodialysis vintage, normalized protein catabolic rate, hemoglobin, serum albumin, creatinine, phosphate, cardiothoracic ratio, logarithmic transformation of intact parathyroid hormone, and high-sensitivity C-reactive protein. Diabetes mellitus was defined as a previous diagnosis of diabetes mellitus or fasting glucose s 126 mg/dl twice subsequently. DISCUSSION The study results demonstrate that blood lead may contribute to an increased risk of cardiovascular-cause, infection-cause, and all-cause mortality in patients on maintenance hemodialysis. If blood lead instead of corrected blood lead is used, similar results are obtained. Moreover, if 19 patients with previous exposure to lead are included, blood lead still contributes to increased risk of mortality in these patients. In the baseline data analysis, patients with high blood lead have higher Kt/V urea values that can be attributed to the higher blood flow rate and dialysate flow rate; however, higher Kt/V is not associated with higher blood lead in these patients after adjusting related variables. High blood lead s and higher parathyroid hormone s may result from hyperparathyroidism-related osteitis fibrosa, which may cause lead stored in bone being released into the bloodstream, resulting in increased blood lead s. 13 Moreover, the study results demonstrated that blood lead is negatively associated with serum ferritin and positively associated with the vintage of hemodialysis and living in an urban area in patients on maintenance hemodialysis. The results are similar to those of previous reports 24,25 on the general population. Patients living in urban areas might have more environmental exposure to lead, and the iron status is negatively related to blood lead because iron may interfere with lead absorption in the gastrointestinal tract. Corrected blood lead s also were associated negatively with hemoglobin and serum albumin. This may result from exposure to lead influencing nutritional status or the effects of corrected hemoglobin for blood lead in these patients. Further study is needed to clarify these observations. Patients on maintenance hemodialysis in this study had a mean blood lead of 11.5 g/dl, which is higher than that observed in the general population in Taiwan (7.7 g/dl) 26 and between that of the general population in Europe (11.4 g/dl) 4 and the United States (2.8 g/dl). 5 These findings are in agreement with those of previous studies performed in patients on maintenance hemodialysis Because hemodialysis centers have very low lead contents ( 2 g/l) in the dialysis water and dialysate, the elevated blood lead in patients on maintenance hemodialysis may be attributed to a complete loss of renal function to excrete lead from the body and the difficulty in removing lead through hemodialysis. 14 Therefore, environmental exposure to lead may increase the blood lead in patients on maintenance hemodialysis. This may explain why blood lead is positively correlated to hemodialysis vintage and not age, 4 as observed in the current study subjects. At the end of the 18-month observation period, 59 patients (6.4%) died. Kaplan Meier survival analysis showed that the high blood lead group had greater mortality than the low blood lead group. The study results demonstrated that, after adjusting for related variables, with the low blood lead (or low corrected blood lead )

8 Lin et al Lead and Mortality in Patients on MHD 357 as the reference, high blood lead (or high corrected blood lead ) was associated with increased HRs for cardiovascular-cause, infection-cause, and all-cause mortality in patients on maintenance hemodialysis. A previous study 14 in diabetic patients on maintenance hemodialysis failed to demonstrate that blood lead or high blood lead is associated with increased HRs for mortality in Cox multivariate analysis. Moreover, the short follow-up period (12-month) and small sample size (n 212) may limit the study results. 14 Another study 15 in patients on chronic peritoneal dialysis failed to demonstrate that high blood lead was associated with increased HRs for cardiovascular-cause and infection-cause mortality in these patients. Moreover, lack of measurement of high-sensitivity C-reactive protein and relatively small sample size (n 315) may weaken the study results. 15 The results of the current study draw a more definite conclusion on the relationship between blood lead s and mortality in patients on dialysis. Moreover, if the diabetic patients on maintenance hemodialysis are excluded from the current study, with the low blood lead group as the reference, high blood lead was still associated with increased HRs for all-cause mortality in nondiabetic patients on maintenance hemodialysis. Because 46% of patients with high blood lead s and 3% patients with low blood lead were from urban areas, living in an urban area may be an important confounder for assessing the relationship between blood lead and mortality in patients on maintenance hemodialysis. However, living in an urban area is not associated with HRs for cardiovascular-cause, infection-cause, and allcause mortality in univariate Cox analysis. Moreover, the high blood lead is still associated with increased HRs for all-cause mortality in countryside patients (n 752) after related variables were adjusted with the low blood lead as the reference. Thus, the blood lead is not simply a marker of an urban life style and may be causal. These study results are also in agreement with those of several epidemiologic studies 1,2 that investigated the general population and demonstrated that blood lead is associated with an increased risk of death from all causes, cardiovascular cause, and cancer cause during the 8.6- and 12-year follow-ups. The reason why blood lead s or corrected blood lead s increased the risk of mortality in patients on maintenance hemodialysis remains unknown. Some studies performed on animals 27,28 have shown that chronic exposure to low-dose lead results in the generation of reactive oxygen species, reduces nitric oxide availability and expression of angiotensin II, and increases blood pressure. 28 Low- exposure to lead also promotes hydroxyl radical generation and lipid peroxidation, 29 enhances vascular reactivity to sympathetic stimulation, and diminishes DNA repair, which may be relevant for rapidly dividing cells in the inflamed arterial wall. 30 Moreover, rats chronically exposed to low s of lead were successfully treated with antioxidants; 28,29 thus, oxidative stress plays a primary role in low s of lead exposure. 31,32 Overproduction of reactive oxygen species in patients on maintenance hemodialysis has been implicated in increased long-term complications, including accelerated inflammation or malnutrition in patients on maintenance hemodialysis, and infection-cause and cardiovascular-cause mortality in these patients. However, further study is needed to clarify the definite pathogenesis. Because of the observational nature of this study, the associations do not necessarily indicate causality. Although blood lead reflects recent exposure to lead, it also is influenced by long-term exposures through efflux of lead from bone stores. 2 Thus, it is unclear whether the adverse health effects of lead that were observed were associated with current or cumulative exposures. A recent investigation 36 showed that blood lead s were significantly and positively associated with bone lead contents in patients with end-stage renal disease after related variables were adjusted. We used the mean of blood lead s by measuring blood lead s twice with a 3-month interval in the current study to avoid bias, but further bone lead assessment is needed to resolve this problem. Moreover, the current study did not enroll incident patients on maintenance hemodialysis, but patients with a different duration of maintenance hemodialysis, which might cause a bias in the hypotheses. However, by adjusting for hemodialysis vintage in each multivariate Cox analysis model, the relationship between blood lead and increased HRs for mortality still exists. Further prospective studies are required to determine whether removal of sources of environmental lead exposure from patients with dialysis therapy or prevention of iron deficiency in these patients improves mortality. Because chelation therapy for low- lead exposure has shown benefits in retarding progressive renal insufficiency, 6-10 it is reasonable to suggest chelation therapy with a low-dose chelating agent for patients on dialysis with high blood lead s 13 if all other interventions are ineffective. CONCLUSIONS This is the first study to demonstrate that blood lead may contribute to increased risk of cardiovascular-cause, infection-cause, and all-cause mortality in patients on maintenance hemodialysis. Our results lend support to the continued efforts to reduce environmental exposure to lead, avoid iron-deficiency status, and explore a possible chelation therapy for patients receiving dialysis. Further studies are required to clarify and confirm these observations. ACKNOWLEDGMENTS The authors thank the staff of 3 hemodialysis centers in the Taipei and Lin-Kou Medical Centers, and the Taoyuan division of Chang Gung Memorial Hospital. References 1. Schober SE, Mirel LB, Graaubard BI, Brody DJ, Flegal KM. Blood lead s and death from all causes, cardiovascular disease, and

9 358 The American Journal of Medicine, Vol 124, No 4, April 2011 cancer: results from the NHANES III mortality study. Environ Health Perspect. 2006;114: Menke A, Muntner P, Batuman V, Silbergeld EK, Guallar E. Blood lead below 0.48 micromol/l (10 microg/dl) and mortality among US adults. Circulation. 2006;114: Nawrot TS, Thijs L, Den Hond EM, Roels HA, Staessen JA. An epidemiological re-appraisal of the association between blood pressure and blood lead: a meta-analysis. J Hum Hypertens. 2002;16: Staessen JA, Lauwerys RR, Buchet JP, et al. The Cadmibel Study Group: Impairment of renal function with increasing blood lead concentrations in the general population. N Engl J Med. 1992;327: Kim R, Rotnitzky A, Sparrow D, Weiss ST, Wager C, Hu H. A longitudinal study of low- lead exposure and impairment of renal function. The Normative Aging Study. JAMA. 1996;275: Lin JL, Tan DT, Hsu KH, Yu CC. Environmental lead exposure and progressive renal insufficiency. Arch Intern Med. 2001;161: Lin JL, Lin-Tan DT, Yu CC, Huang YY, Li KL. Environmental exposure to lead and progressive diabetic nephropathy in patients with type II diabetes. Kidney Int. 2006;69: Lin JL, Lin-Tan DT, Hsu KH, Yu CC. Environmental lead exposure and progression of chronic renal diseases in patients without diabetes. New Engl J Med. 2003;348: Lin JL, Ho HH, Yu CC. Chelation therapy for patients with elevated body lead burden and progressive renal insufficiency. A randomized, clinical trial. Ann Intern Med. 1999;130: Lin-Tan DT, Lin JL, Yen TH, Chen KH, Huang YL. Long-term outcome of repeated lead chelation therapy in progressive non-diabetic chronic kidney diseases. Nephrol Dial Transplant. 2007;22: Krachler M, Wirnsberger GH. Long-term changes of plasma trace element concentrations in chronic hemodialysis patients. Blood Purif. 2000;18: Colleoni N, Arrigo G, Gandini E, Corigliano C, D Amico G. Blood lead in hemodialysis patients. Am J Nephrol. 1993;13: Kessler M, Durand PY, Huu TC, et al. Mobilization of lead from bone in end-stage renal failure patients with secondary hyperparathyroidism. Nephrol Dial Transplant. 1999;14: Lin JL, Lin-Tan DT, Yen TH, et al. Blood lead s, malnutrition, inflammation, and mortality in patients with diabetes treated by longterm hemodialysis. Am J Kidney Dis. 2008;51: Lin JL, Lin-Tan DT, Chen KH, et al. Blood lead s association with 18-month all-cause mortality in patients with chronic peritoneal dialysis. Nephrol Dial Transplant. 2010;25: Huang CC, Cheng KF, Wu HD. Survival analysis: comparing peritoneal dialysis and hemodialysis in Taiwan. Perit Dial Int. 2008; 28(Suppl 3):S15-S US Renal Data System. Patient mortality and survival. Am J Kidney Dis. 1998;32(Suppl 1):S69-S Lin JL, Tan DT, Ho HH, Yu CC. Environmental lead exposure and urate excretion in the general population. Am J Med. 2002;113: Sargent JA. Control of dialysis by a single-pool urea model: The National Cooperative Dialysis Study. Kidney Int. 1983;13(Suppl):S19- S Daugirdas JT. The post: pre-dialysis plasma urea nitrogen ratio to estimate Kt/V and npcr: mathematical modeling. Int J Artif Organs. 1989;12: Chen KH, Lin-Tan DT, Huang WH, et al. Cardiothoracic ratio, malnutrition, inflammation, and two-year mortality in non-diabetic patients on maintenance hemodialysis. Kidney Blood Press Res. 2008; 31: Lin-Tan DT, Lin JL, Wang LH, et al. Fasting glucose s in predicting 1-year all-cause mortality in patients who do not have diabetes and are on maintenance hemodialysis. J Am Soc Nephrol. 2007;18: Kalantar-Zadeh K. Recent advances in understanding the malnutritioninflammation-cachexia syndrome in chronic kidney disease patients: what is next? Semin Dial. 2005;18: Schwela D. Air pollution and health in urban areas. Rev Environ Health. 2000;15: Goyer RA. Lead toxicity: current concerns. Environ Health Perspect. 1993;100: Liou SH, Wu TN, Chiang HC, et al. Three-year survey of blood lead s in 8828 Taiwanese adults. Int Arch Occup Environ Health. 1996;69: Rodriguez-Iturbe B, Sindhu RK, Quiroz Y, Vaziri ND. Chronic exposure to low doses of lead results in renal infiltration of immune cells, NF-kappaB activation, and overexpression of tubulointerstitial angiotensin II. Antioxid Redox Signal. 2005;7: Vaziri ND. Mechanisms of lead-induced hypertension and cardiovascular disease. Am J Physiol Heart Circ Physiol. 2008;295:H454-H Ding Y, Gonick HC, Vaziri ND. Lead promotes hydroxyl radical generation and lipid peroxidation in cultured aortic endothelial cells. Am J Hypertens. 2000;13(5 Pt 1): Nawrot TS, Staessen JA. Low- environmental exposure to lead unmasked as silent killer. Circulation. 2006;114: Aykin-Burns N, Franklin EA, Ercal N. Effects of N-acetylcysteine on lead-exposed PC-12 cells. Arch Environ Contam Toxicol. 2005;49: Ding Y, Vaziri ND, Gonick HC. Lead induced hypertension. II. Response to sequential infusions of L-arginine, superoxide dismutase, and nitroprusside. Environ Res. 1998;76: Morena M, Delbosc S, Dupuy AM, Canaud B, Cristol JP. Overproduction of reactive oxygen species in end-stage renal disease patients: a potential component of hemodialysis-associated inflammation. Hemodial Int. 2005;9: Yao Q, Pecoits-Filho R, Lindholm B, Stenvinkel P. Traditional and non-traditional risk factors as contributors to atherosclerotic cardiovascular disease in end-stage renal disease. Scand J Urol Nephrol. 2004;38: Lin TH, Chen JG, Liaw JM, Juang JG. Trace elements and lipid peroxidation in uremic patients on hemodialysis. Biol Trace Elem Res. 1996;51: Muntner P, Menke A, Batuman V, Rabito FA, He J, Todd AC. Association of tibia lead and blood lead with end-stage renal disease: a pilot study of African-Americans. Environ Res. 2007; 104: