NDT Advance Access originally published online on December 15, 2006
Nephrology Dialysis Transplantation 2007 22(3):891-898; doi:10.1093/ndt/gfl689
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The association between recipient alcohol dependency and long-term graft and recipient survival
1Department of Biomedical Informatics, University of Utah, 2Division of Nephrology and Hypertension, University of Utah School of Medicine, Salt Lake City, UT, 3Department of Internal Medicine, University of Nevada Las Vegas, Las Vegas, NV, USA, 4Tel Aviv University, Tel Aviv, Israel, 5Division of Undergraduate Studies and University Writing Program, University of Utah and 6Veterans Affairs Salt Lake City Healthcare System, Salt Lake City, UT, USA
Correspondence and offprint requests to: Alexander S. Goldfarb-Rumyantzev, MD, PhD, Division of Nephrology and Hypertension, University of Utah Health Sciences Center, 85 North Medical Drive, East Room 201, Salt Lake City, UT 84112, USA. Email: alex.goldfarb{at}hsc.utah.edu
 |
Abstract |
|---|
 Top Abstract BackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
Background. The causative role of alcohol consumption in renal disease is controversial, and its effect on renal graft and recipient survival has not been previously studied.
Methods. We analysed the association between pre-transplant [at the time of end-stage renal disease (ESRD) onset] alcohol dependency and renal graft and recipient survival. The United States Renal Data System (USRDS) records of kidney transplant recipients 18 years or older transplanted between 1 January 1995 and 31 December 2002 were examined. We used Kaplan–Meier analysis and Cox regression models adjusted for covariates to analyse the association between pre-transplant alcohol dependency and graft and recipient survival.
Results. In an entire study cohort of 60 523, we identified 425 patients with a history of alcohol dependency. Using Cox models, alcohol dependency was found to be associated with increased risk of death-censored graft failure [hazard ratio (HR) 1.38, P < 0.05] and increased risk of transplant recipient death (HR 1.56, P < 0.001). Subgroup analysis demonstrated an association of alcohol-dependency with recipient survival and death-censored graft survival in males (but not in females), and in both white and non-white racial subgroups.
Conclusions. We concluded that alcohol dependency at the time of ESRD onset is a risk factor for renal graft failure and recipient death.
Keywords: alcohol dependency; graft survival; kidney transplantation; outcome; prediction; public policy; recipient survival; renal transplantation
|
Background |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
Renal transplantation is the modality of choice for treatment of patients with end-stage renal disease (ESRD). Although the rates of renal graft and recipient survival have improved over the past several years [1], further extending the allograft and recipient survival remains an important problem. Despite a relatively high prevalence of alcohol consumption in the transplant population [2] and an increasing awareness of renal complications associated with illicit drugs [3], the potential impact of alcohol on renal function remains a controversial issue [4]. In the general non-transplant population, several studies have demonstrated that moderate consumption (1–3 drinks per day) of alcohol has a beneficial effect in the prevention of coronary artery disease, stroke and diabetes mellitus [5–7].
Studies on the renal effects of alcohol consumption have reported various results. Experimentally, alcohol may cause reduced renal function, interstitial oedema and renal hypertrophy in alcohol-fed animals [8]. However, a large prospective trial found no adverse outcome on renal function with moderate alcohol intake in healthy women [9]. Another cohort study showed that healthy men who consumed at least seven drinks per week were not at risk of decline in renal function. Instead, the authors found that men with moderate alcohol consumption had a 29% reduction in risk of elevated creatinine levels in a 14-year follow-up period [10]. Similarly, Chung et al. [4] demonstrated that chronic alcohol users and ex-users had higher estimated creatinine clearances and glomerular filtration rates compared with non-drinking controls.
The findings of some of these studies suggest that moderate alcohol consumption might have a favourable outcome in individuals with normal renal function. But in renal transplant recipients and those with pre-existing renal disease, its effect is still not clearly understood. A recently published study of Swiss patients reported that 52.4% of renal transplant recipients consumed alcohol at least once a week, and that 1.5% of them were drinking at a moderate health risk level per World Health Organization classification [2]. While there is some evidence that alcohol consumption is associated with late-onset renal failure in patients who underwent liver transplantation due to hepatitis C [11], there is no information describing the effect of alcohol consumption on renal function in other organ transplant recipients.
With these previous studies in mind, we conducted a retrospective transplant registry data analysis to estimate the effect of pre-transplant alcohol dependency on renal graft and recipient survival.
|
Methods |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
Data set
The data collected by the United States Renal Data System (USRDS), including transplant-related data collected by the United Network for Organ Sharing (UNOS) were used for the analysis. The USRDS records of post-kidney transplant patients 18-years old or older transplanted during the period of 1 January, 1995, and 31 December, 2002, were examined. Only records with non-missing information regarding the outcome and primary variable of interest were used in the analysis (n = 60 523).
Outcome variables
We analysed the data in regard to two clinical outcomes: (i) graft survival (defined as the time between the most recent kidney transplant procedure and the date of return to dialysis, re-transplant, or censor) and (ii) recipient survival (defined as the time between the most recent kidney transplant procedure and recipient death or censor). Graft failure date and recipient death date were derived from the following USRDS files: TXUNOS, RXHIST60, PATIENT. The graft survival analysis did not include death with functioning graft (i.e. death-censored graft survival) based on USRDS definition. In cases where this information was missing, we assumed that if the graft failure date and the date of patient death were equal the patient died with a functioning graft. Our intention was to avoid an over-reporting bias, which means mistakenly to include those who died on the same day as the graft failure was reported, when in fact the graft dysfunction occurred independent of patient death. Therefore, the above rule was not applied to the cases when the cause of death was coded by ICD-9 as one of the following: 3200 (graft failure: primary failure), 3201 (graft failure: rejection), 3202 (graft failure: technical), 3299 (graft failure: other) or 3903 (miscellaneous: renal failure). Allograft outcome was censored at the earliest of the following events: loss to follow-up, patient death or the study completion date (30 June 2006), and was analysed as days-to-graft-failure or censor. Patient follow-up was censored at the earlier time of loss to follow-up or study completion date, and was analysed as time-to-recipient-death or censor.
Independent variables
Primary variable of interest
The primary variable of interest for this study was alcohol dependency recorded at first ESRD service. The information was obtained from The Centers for Medicare and Medicaid Services (CMS). Alcohol dependence is listed under comorbid conditions in the Chronic Renal Disease Medical Evidence Form (CMS-2728). Physician, head nurse, or social worker involved in the patient's treatment of renal disease completes the CMS-2728. This information was collected from the Medical Evidence file in the USRDS database, where alcohol dependency is described by a single categorical (yes, no, unknown) variable ALCOH. The frequency of alcohol-dependent recipients is presented in Table 1.
|
Covariates
The Cox models were adjusted for covariates believed to be potential confounding factors to the primary variable of interest and the outcome (i.e. information potentially affecting the outcome and the primary variable of interest). The decision concerning which covariates to include in the final models was based primarily on the known associations between variables that could cause confounding. Based on our previous work [12,13] and that of others [14,15], all models were adjusted for the following variables.
Recipient variables
Age; gender; height; weight; race; total number of transplants; cause of ESRD (diabetes mellitus (DM); hypertension (HTN); glomerulonephritis (GN) or other/unknown); history of diabetes; history of hypertension; and other comorbidity information (represented as a comorbidity index defined below); duration of pre-transplantation ESRD; primary source of pay for medical service; citizenship; and education level.
Donor variables
Donor type (living or deceased); race.
Transplantation procedure parameters
Cold storage time, peak panel reactive antibody level, procedure type [simultaneous kidney/pancreas transplantation (SPK) or not].
The comorbidity index for recipients was calculated based on the Charlson index described elsewhere [16], and previously validated in the kidney transplant population [17]. Only the data available in the data set were used to calculate the Charlson index: myocardial infarction, congestive heart failure, ischaemic heart disease, peripheral vascular disease, cerebrovascular disease, chronic obstructive pulmonary disease, cancer and diabetes. For each decade over 40 years of age, an additional point was added to the index. This somewhat abbreviated version of the Charlson index was previously validated by our group in renal transplant recipients [18]. Both the Charlson index and history of diabetes were included in the Cox models because diabetic status is a significant predictor of the outcomes, and its effects cannot be completely explained by the Charlson index. The lack of colinearity between diabetic status and the Charlson index was demonstrated by the correlation coefficient 0.09.
Statistical analysis
Unrealistic values of the independent variables used in the study were eliminated. The heights and weights of recipients (only adults were included in this study) and donors were based on the acceptable ranges: height (120–275 cm), weight (23–180 kg). The entire data set and sub-groups of patients were analysed. Kaplan–Meier graphs were used to display hypothesized relationships, and Cox regression models were used to adjust for the covariates and analyse the survival information. Data were analysed using SAS version 9.1 (SAS Institute, Cary, NC). The subgroup analysis was conducted by stratifying the entire patient population by gender, education level (college vs non-college), race (white vs non-white), procedure type (SPK vs non-SPK) and donor type (living vs deceased).
Cold storage time variable was missing information and was imputed by using a multiple imputation technique. Multiple imputation is the method of substituting missing values using a set of plausible values and involves the following steps: (i) creation of several simulated complete versions of the data set; (ii) analyses of each new data set; and (iii) pooling results from the analyses [19]. The tools for multiple imputation used in this analysis are available in SAS version 9 (procedures PROC MI and PROC MIANALYZE).
|
Results |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
Baseline characteristics
The baseline characteristics of the entire study population compared with the alcohol-dependent and non-dependent groups are represented in Table 1. The study population (n = 60 523) with a mean age of 44.6 ± 14.9 years included 39.7% females, 20.8% African Americans and 70.0% Whites. The mean recipient ESRD duration was 32.1 ± 40.5 months, and 36.8% of the recipients had living donors. In addition, 33.2% of the recipients had a history of diabetes, and 72.5% had a history of hypertension. The Charlson comorbidity index was 2 ± 1.7. One of three patients was college educated and 47.4% had private insurance. Only 0.7% of the recipients had alcohol dependency at the time of the first ESRD service.
Characteristics of alcohol-dependent population
There were 425 recipients who were alcohol dependent. This population was 3.9 years older than the non-alcohol-dependent group (P < 0.001). The racial and the sex ratio comparison between the alcohol-dependent and non-dependent groups showed some differences. We did not find any Asians with alcohol-dependency. Native Americans were represented three times more in the alcohol-dependent group than the non- alcohol-dependent group. African Americans were also more likely to be alcohol dependent than whites. Men were three times more likely to be alcohol-dependent than women. The peak PRA levels were lower among the alcohol-dependent population than among the non-alcohol-dependent group. The alcohol-dependent group was more likely to receive a kidney with longer cold-ischaemia time compared with the non-alcohol-dependent group. Members of this group were also less likely to get a kidney from a living donor. The alcohol-dependent recipients were less likely to be diabetic and more likely to be hypertensive than the non-alcohol-dependent recipients. Alcohol-dependent recipients had a higher Charlson comorbidity index compared with the non-alcohol-dependent group.
Survival analysis
Kaplan–Meier survival analysis suggests that death-censored graft survival was worse for the alcohol-dependent recipients (P < 0.01, Figure 1). Similarly, in this unadjusted analysis, alcohol-dependent recipients survive for shorter time periods than the non-alcohol-dependent recipients (P < 0.001, Figure 2). The results of the Cox modelling for the survival analysis for the entire study population are presented in Table 2. In the Cox model, adjusting for potential confounding variables, alcohol dependency was found to be associated with an increased risk of graft failure (HR 1.38, P < 0.05). Similarly, alcohol dependency at the time of the first ESRD service was associated with increased risk of recipient death (HR 1.56, P < 0.001).
|
2 = 7.59 and P < 0.01). The duration of the follow-up was set at 6 years due to the fact that no grafts in the alcohol-dependent group survived beyond 6 years.
|
|
Subgroup analysis
The subgroup analysis results are presented in Table 3. Recipient history of alcohol dependence still plays a significant role in death-censored graft survival when stratified by gender. For males, alcohol dependency was associated with HR for graft failure of 1.56 (P < 0.01), whereas females were not analysed separately due to sample size restrictions. Separate analysis by race yielded significant results (HR 1.64, P < 0.005 for white vs HR 1.14, P < 0.001 for non-white). The effect of alcohol dependency is no longer significant when we stratified the patients by education level, transplant procedure type or donor type.
|
The relationship between recipient history of alcohol dependency and patient survival is significant for males (HR 1.58, P < 0.001), but not females. Similarly, history of alcohol dependency plays a significant role on recipient survival when the analysis was stratified by race (HR 1.70, P < 0.001 for white vs HR 1.39, P < 0.001 for non-white). Significant results were also observed for patients without college education (HR 1.53, P < 0.001), but not for those with college education. Alcohol dependency was also associated with shorter recipient survival in the subgroup of patients who were transplanted with kidney alone (HR 1.57, P < 0.001), but not for those with SPK transplant. This association was also present in the recipients of deceased donor grafts (HR 1.55, P < 0.001), but not in the recipients of living donor grafts.
|
Discussion |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
In the present study, we analysed the association between alcohol dependency and renal graft and recipient survival. To the best of our knowledge, this report is the first study evaluating the role of recipient alcohol dependency in kidney transplant outcome. In our previous study, we evaluated the role of alcohol dependency in donors. We demonstrated that donor's alcohol dependency did not have a significant adverse impact on graft or recipient survival [20]. In this analysis, we showed that a history of recipient alcohol dependency at the time of ESRD onset has a negative effect on both graft and recipient survival. Our analysis suggests that alcohol-dependent recipients have a shorter survival time than non-alcohol-dependent recipients, and that their grafts fail earlier as well. The subgroup analysis corroborates these results.
While we found no literature specifically addressing this subject, there are studies related to ours, which may be potentially helpful in explaining the mechanism of the observed association. Several observational studies have found a J-shaped association between alcohol consumption and hypertension, where three or more drinks per day were associated with increased prevalence of hypertension, while two or fewer drinks a day were associated with lower prevalence of hypertension [21]. These findings were also confirmed in large prospective studies [22,23]. Interestingly, the Prevention and Treatment of Hypertension Study reported that reducing the daily intake of alcohol from 432 g per week to 208 g per week did not result in a significant reduction in blood pressure compared with the control group [24]. While the relationship between hypertension and graft survival is a plausible, but not necessarily direct one, there are reports addressing the effect of alcohol consumption on kidney function directly. A large Australian study found that two or fewer drinks per day of alcohol were associated with higher serum creatinine levels and higher blood pressure than their non-drinking controls. However, in the same study consuming three or more drinks per day was associated with lower serum creatinine compared with non-drinkers [25]. The results of this analysis should be interpreted cautiously, as they might be affected by some biases and confounding. In particular, lower serum creatinine might be a reflection of malnutrition status rather than better kidney function. In addition, reverse causality, reporting bias and confounding by unaccounted factors are also possible. In another population-based case-control study, authors reported the J-curve association between the amount of alcohol consumed and risk of ESRD in unadjusted analysis. Interestingly, this association disappeared after exclusion of persons who had ever consumed home-distilled whiskey as well as model adjustment for age, race, sex, income, history of hypertension, history of diabetes mellitus, use of acetaminophen, use of opiates and cigarette smoking. In this multivariate analysis of patients who had started treatment for ESRD and randomly selected controls from the general population, the authors demonstrated that consumption of more than two drinks of alcohol, on average, per day was associated with a 4-fold increased risk of ESRD compared with non-ESRD controls [26]. Another observational study of liver transplant recipients found that alcohol consumption was associated with a 10-fold increased risk of late-onset renal failure, defined as serum creatinine more than 2 mg/dl developing 6 months after transplantation [11]. These results are mostly consistent with our observations presented in this report.
The mechanism by which alcohol consumption can cause renal disease is not clear. Alcohol-dependent patients may have lifestyle habits that adversely affect survival, and they may also not be compliant with their post-transplant treatment and follow-up care. Both factors may lead to poor graft and recipient survival. Paukov and Erokhin [27] have shown that persons who had abused alcohol had progressing alteration of microcirculatory bed, fat degeneration of parenchymatous organs, and atrophic and sclerotic processes, primarily in the liver, lungs, heart and brain [27]. In addition, as discussed above, chronic heavy alcohol consumption may be associated with hypertension, which in turn can lead to renal disease. There are also potential immunological mechanisms pertinent to allograft survival. Alcohol is known to modulate the immune system in a complex manner. For example, alcoholics have increased circulating IgA levels regardless of their liver status, which could be related to antibody production, as well as dysregulation of cytokine production [28]. The effect of alcohol on the immune responses varies with duration of consumption, whether exposure was acute or chronic, and the blood level of alcohol [29]. While acute alcohol ingestion inhibits proinflammatory cell activation [30], chronic alcohol use in humans has the opposite effect [29,31]. Prolonged alcohol use activates monocytes and macrophages, resulting in massive increases in proinflammatory cytokines including tumour necrosis factor-
, interleukin (IL)-1, IL-6 and IL-8 [31]. In addition, chronic alcohol ingestion causes severe oxidative stress and depletion of the antioxidant glutathione [32]. The profibrotic cytokine transforming growth factor (TGF-ß) has been associated with the development of chronic allograft nephropathy, which is characterized by renal graft fibrosis leading to early loss of function [33]. In the animal model, chronic alcohol ingestion has been demonstrated to increase the expression of TGF-ß [34].
This study is an analysis of the existing data registry, which has limitations (i.e. potential quality of data concerns, limitations by the data available in the registry) as well as advantages (i.e. larger sample size and increased statistical power, longer follow-up for long-term outcome analysis). Due to the large sample size, the small and clinically non-significant associations might still demonstrate statistical significance in the analysis.
Some of the important limitations of this study should be realized by the reader. One of them concerns the problems associated with the reporting of alcohol dependency, which might impact the reliability and sensitivity in detecting alcohol abuse. Alcohol dependency information obtained from self-reported questionnaires is always fraught with under-reporting bias, and research shows that this problem is even more pronounced in the case of heavy drinkers [35]. Alcohol dependency is difficult to quantify, and subjects might be unwilling to share certain information. In addition, commonly, respondents do not take into consideration infrequent binge drinking episodes, which nevertheless may present potential health risks. Finally, people who once were heavy alcohol consumers may have cut down or even stopped alcohol consumption after they developed serious health conditions, such as ESRD. Such patients may report their current alcohol consumption status rather than their previous high level of alcohol consumption. Moreover, alcohol dependency was stated at first ESRD service. Since the duration of the pre-transplant period can reach months or years, the status of the patient in respect to alcohol consumption might have changed at the time of transplantation. Unfortunately, the information related to alcohol dependency at the time of transplantation was not available to us. That, however, does not change the conclusion of the study, where we specifically state that alcohol dependency at the time of onset of ESRD (rather that at the time of transplant) is associated with adverse transplant outcomes. All these factors may introduce misclassification and under-reporting bias. Misclassification bias is difficult to address in the retrospective study, and remains one of the shortcomings of the data registry analyses. However, these issues of potential under-reporting and misclassification would weaken the association between alcohol consumption and clinical outcomes, so the observed positive associations presented here should still be valid. On the other hand, one might argue, that the over-estimation of the effect is a possibility too, if the alcohol dependence is simply a label for some other factors confounding the results, and if the population is super-selected based on these other factors. However, this possibility seems to be less likely to us, considering the data were analysed in multivariate analysis adjusted for all the important predictors of the graft outcome.
Another related issue is concerned with the fact that the alcohol-dependent recipients represent a very small fraction of the whole data set (425 recipients with a history of alcohol dependency out of 60 523 total study population). However, even though the number of patients in the groups is so disproportionate, we believe that this issue does not present a study design problem, as generally, in epidemiology research the absolute number of subjects in the group is much more important than the way the primary variable of interest divides the data set.
Second, a retrospective analysis does not provide explanations for the mechanism of the observed associations, and is susceptible to confounding by the factors which were not included in the multivariate models (e.g., depression, poverty and additional indicators of the socio-economic status: annual income, employment status, geographic location, IQ level, marital status and nicotine addiction). Analysis of the role of these variables may present a new and exciting opportunity for future research. Our models were adjusted for several socio-economic status variables, and in addition, analysis was stratified by education level and racial characteristics. The information regarding concurrent smoking or underlying depression is lacking in the data set, yet these factors, if unaccounted for, might confound the analysis.
Third, reverse causality, described elsewhere [36], where poor health may cause alcohol dependency but not vice versa, as suggested by our results, is also possible. However, as the alcohol-dependency information was collected at the time of ESRD onset, we assume that the subsequent selection of patients for the transplant list should have eliminated those in marginal health.
Fourth, there is no information on the amount and frequency of alcohol consumption in the USRDS and UNOS databases. This information could be important, given the postulated J-shaped association between the amount of alcohol consumed and renal disease and hypertension in the general population. The dose-dependent association between alcohol consumption and the transplant outcome cannot be established based on the data available for analysis, and may potentially present the opportunity for future research.
In conclusion, a history of alcohol dependency at the time of onset of ESRD is associated with shorter graft and patient survival. Further investigations are necessary to confirm these findings and to better understand the mechanism of observed association.
|
Acknowledgement |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
This study was supported in part by the Dialysis Research Foundation (Ogden, UT). The data reported here have been supplied by the USRDS. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as official policy or interpretation of the US government.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionAcknowledgementReferences |
|---|
- Understanding Transplant Statistics. Transplant News (2002/2003).
- Fierz K, Steiger J, Denhaerynck K, Dobbels F, Bock A, De Geest S. (2006) Prevalence, severity and correlates of alcohol use in adult renal transplant recipients. Clin Transplant 20:171–178.[CrossRef][Web of Science][Medline]
- Crowe AV, Howse M, Bell GM, Henry JA. (2000) Substance abuse and the kidney. Qjm 93:147–152.
[Abstract/Free Full Text] - Chung FM, Yang YH, Shieh TY, Shin SJ, Tsai JC, Lee YJ. (2005) Effect of alcohol consumption on estimated glomerular filtration rate and creatinine clearance rate. Nephrol Dial Transplant 20:1610–1616.
[Abstract/Free Full Text] - Gaziano JM, Buring JE, Breslow JL, et al. (1993) Moderate alcohol intake, increased levels of high-density lipoprotein and its subfractions, and decreased risk of myocardial infarction. N Engl J Med 329:1829–1834.
[Abstract/Free Full Text] - Berger K, Ajani UA, Kase CS, et al. (1999) Light-to-moderate alcohol consumption and risk of stroke among U.S. male physicians. N Engl J Med 341:1557–1564.
[Abstract/Free Full Text] - Ajani UA, Hennekens CH, Spelsberg A, Manson JE. (2000) Alcohol consumption and risk of type 2 diabetes mellitus among US male physicians. Arch Intern Med 160:1025–1030.
[Abstract/Free Full Text] - Van Thiel DH, Gavaler JS, Little JM, Lester R. (1977) Alcohol. its effect on the kidney. Metabolism 26:857–866.[CrossRef][Web of Science][Medline]
- Knight ESM, Rimm E, Hankinson S, Curhan G. Moderated alcohol intake and renal function decline in women: a prospective study. Nephrol Dial Transpl 18:1549–1554.
- Schaeffner ES, Kurth T, de Jong PE, Glynn RJ, Buring JE, Gaziano JM. (2005) Alcohol consumption and the risk of renal dysfunction in apparently healthy men. Arch Intern Med 165:1048–1053.
[Abstract/Free Full Text] - Gayowski T, Singh N, Keyes L, et al. (2000) Late-onset renal failure after liver transplantation: role of posttransplant alcohol use. Transplantation 69:383–388.[CrossRef][Web of Science][Medline]
- Goldfarb-Rumyantzev A, Hurdle JF, Scandling J, et al. (2005) Duration of end-stage renal disease and kidney transplant outcome. Nephrol Dial Transplant 20:167–175.
[Abstract/Free Full Text] - Goldfarb-Rumyantzev AS, Scandling JD, Pappas L, Smout RJ, Horn S. (2003) Prediction of 3-yr cadaveric graft survival based on pre-transplant variables in a large national dataset. Clin Transplant 17:485–497.[CrossRef][Web of Science][Medline]
- Gjertson DW. (2000) Determinants of long-term survival of adult kidney transplants: a 1999 UNOS update. Clin Transplants 1999(UCLA Immunogenetics Center, Los Angeles) pp. 341–352.
- Meier-Kriesche H, Port FK, Ojo AO, et al. (2001) Deleterious effect of waiting time on renal transplant outcome. Transplant Proc 33:1204–1206.[CrossRef][Web of Science][Medline]
- Charlson ME, Pompei P, Ales KL, MacKenzie CR. (1987) A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40:373–383.[CrossRef][Web of Science][Medline]
- Jassal SV, Schaubel DE, Fenton SS. (2005) Baseline comorbidity in kidney transplant recipients: a comparison of comorbidity indices. Am J Kidney Dis 46:136–142.[CrossRef][Web of Science][Medline]
- Lin SJ, Koford JK, Baird BC, et al. (2006) The association between length of post-kidney transplant hospitalization and long-term graft and recipient survival. Clin Transplant 20:245–252.[CrossRef][Web of Science][Medline]
- McCleary L. (2002) Using multiple imputation for analysis of incomplete data in clinical research. Nurs Res 51:339–343.[CrossRef][Web of Science][Medline]
- Lin SJ, Koford JK, Baird BC, et al. (2005) Effect of donors’ intravenous drug use, cigarette smoking, and alcohol dependence on kidney transplant outcome. Transplantation 80:482–486.[CrossRef][Web of Science][Medline]
- MacMahon S. (1987) Alcohol consumption and hypertension. Hypertension 9:111–121.
[Abstract/Free Full Text] - Tsuruta M, Adachi H, Hirai Y, Fujiura Y, Imaizumi T. (2000) Association between alcohol intake and development of hypertension in Japanese normotensive men: 12-year follow-up study. Am J Hypertens 13:482–487.[CrossRef][Web of Science][Medline]
- Witteman JC, Willett WC, Stampfer MJ, et al. (1990) Relation of moderate alcohol consumption and risk of systemic hypertension in women. Am J Cardiol 65:633–637.[CrossRef][Web of Science][Medline]
- Bulpitt CJ and Shipley MJ. (1999) Failure of alcohol reduction to lower blood pressure in the PATHS trial. Prevention and treatment of hypertension Study. Arch Intern Med 159:195–196.
[Free Full Text] - Savdie E, Grosslight GM, Adena MA. (1984) Relation of alcohol and cigarette consumption to blood pressure and serum creatinine levels. J Chronic Dis 37:617–623.[CrossRef][Web of Science][Medline]
- Perneger TV, Whelton PK, Puddey IB, Klag MJ. (1999) Risk of end-stage renal disease associated with alcohol consumption. Am J Epidemiol 150:1275–1281.
[Abstract/Free Full Text] - Paukov VS and Erokhin Iu A. (2004) Pathologic anatomy of hard drinking and alcoholism. Arkh Patol 66:3–9.[Medline]
- Sheron N. (1994) Alcoholic liver damage—toxicity, autoimmunity and allergy. Clin Exp Allergy 24:503–507.[CrossRef][Web of Science][Medline]
- Crews FT, Bechara R, Brown LA, et al. (2006) Cytokines and alcohol. Alcohol Clin Exp Res 30:720–730.[CrossRef][Web of Science][Medline]
- Szabo G, Mandrekar P, Catalano D. (1995) Inhibition of superantigen-induced T cell proliferation and monocyte IL-1 beta, TNF-alpha, and IL-6 production by acute ethanol treatment. J Leukoc Biol 58:342–350.[Abstract]
- McClain CJ, Barve S, Deaciuc I, Kugelmas M, Hill D. (1999) Cytokines in alcoholic liver disease. Semin Liver Dis 19:205–219.[Web of Science][Medline]
- Moss M, Guidot DM, Wong-Lambertina M, Ten Hoor T, Perez RL, Brown LA. (2000) The effects of chronic alcohol abuse on pulmonary glutathione homeostasis. Am J Respir Crit Care Med 161:414–419.
[Abstract/Free Full Text] - Jain S, Furness PN, Nicholson ML. (2000) The role of transforming growth factor beta in chronic renal allograft nephropathy. Transplantation 69:1759–1766.[CrossRef][Web of Science][Medline]
- Bechara RI, Brown LA, Roman J, Joshi PC, Guidot DM. (2004) Transforming growth factor beta1 expression and activation is increased in the alcoholic rat lung. Am J Respir Crit Care Med 170:188–194.
[Abstract/Free Full Text] - Hesselbrock M, Babor TF, Hesselbrock V, Meyer RE, Workman K. (1983) "Never believe an alcoholic"? On the validity of self-report measures of alcohol dependence and related constructs. Int J Addict 18:593–609.[Web of Science][Medline]
- Greenberg JA. (2002) Hypothesis–the J-shaped follow-up relation between mortality risk and disease risk-factor is due to statistical confounding. Med Hypotheses 59:568–576.[CrossRef][Web of Science][Medline]
Accepted in revised form: 25.10.06
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||