Decreased body mass index as an independent risk factor for developing chronic kidney disease.

BACKGROUND: Obesity and metabolic syndrome are risk factors for the development of chronic kidney disease (CKD). Few studies have examined the effect of change in body mass index (DeltaBMI) on CKD incidence in a general screening setting. METHODS: Subjects of this study were screenees that participated in the screening program of the Okinawa General Health Maintenance Association in 1993 and 2003 in Okinawa, Japan. Using identification number, birth date, sex, and other recorded identifiers, we identified 33,389 subjects among the 1993 screening participants (N = 143,948) who also participated in the 2003 screening. CKD was defined as estimated glomerular filtration rate <60 ml/min/1.73 m(2), according to the modification of diet in renal disease study equation. Obesity was defined as BMI > or = 25 kg/m(2). RESULTS: CKD prevalence was 13.8% in 1993 and 22.4% in 2003. The incidence of developing CKD in 10 years was 15.5%. The effect of DeltaBMI on CKD incidence was evaluated after considering other confounding factors such as age, sex, blood pressure, BMI, fasting plasma glucose, and proteinuria. Median DeltaBMI was 1.0%. The adjusted odds ratio (95% CI) for the effect of DeltaBMI on CKD incidence was 1.111 (1.026-1.204, P < 0.01; entire study population), 1.271 (1.116-1.448, P = 0.0030; men), and 1.030 (0.931-1.139, NS; women), when DeltaBMI > or = 1% was taken as a reference. DeltaBMI was an independent predictor of CKD incidence. CONCLUSIONS: The present results suggest that there was an inverse relationship between DeltaBMI and CKD incidence among screened subjects. The reasons for this observation are not clear, but careful follow-up for DeltaBMI is necessary, particularly in obese men with proteinuria.

Metabolic syndrome and risk of developing chronic kidney disease in Japanese adults.

Metabolic syndrome is a risk factor for the development of cardiovascular disease. Few prospective studies, however, have examined metabolic syndrome as a risk factor for chronic kidney disease (CKD) in an Asian population. We studied the occurrence of CKD in 6,371 subjects without CKD or diabetes mellitus at baseline 1997 through 2002 in Okinawa, Japan. CKD was defined as dipstick-positive proteinuria (>or=1+) or a low estimated glomerular filtration rate (<60 mL/min/1.73 m2). Metabolic syndrome was defined according to the modified criteria of the Adult Treatment Panel III in which body mass index (>or=25 kg/m2) was substituted for the waist circumference measurement. Logistic analysis was used to analyze the effect of metabolic syndrome on the development of CKD. During the 5-year follow-up, 369 (5.7%) participants developed CKD. After adjusting for age, sex, current cigarette smoking and alcohol drinking habits at baseline, the relative risk of developing CKD was 1.86 (95% confidence interval: 1.43-2.41, p<0.0001) in subjects with metabolic syndrome. Compared with those without metabolic syndrome risk components, the adjusted relative risk (95% confidence interval) was 1.49 (1.10-2.01), 1.89 (1.38-2.59), and 2.65 (1.19-3.68) in those with 1, 2, or >or=3 metabolic syndrome risk components, respectively. Metabolic syndrome is a significant risk factor for the development of CKD in the Japanese population. Detection and treatment of metabolic syndrome should be stressed as a strategy to prevent CKD.

Relationship between dyslipidemia and the risk of developing end-stage renal disease in a screened cohort.

BACKGROUND: Disturbances in lipid metabolism are often observed in patients with renal failure and could be a risk factor for end-stage renal disease (ESRD). However, few studies have examined abnormal lipid metabolism as a risk factor for the development of ESRD in the general population. METHODS: We examined the cumulative incidence of ESRD based on the results of a community-based mass screening in Okinawa, Japan, which was conducted in 1993 by the Okinawa General Health Maintenance Association. Screenees who developed ESRD by the end of 2000 were identified through the Okinawa Dialysis Study registry. RESULTS: Total cholesterol (TC) data were available for 133,338 (92.6%) of the total 143,948 screenees) and triglyceride (TG) data were available for 132,094 (91.8%). Dyslipidemia was defined as TC > or = 220 mg/dl or TG > or = 150 mg/dl. The cumulative incidences of ESRD, per 1000 screenees, were 1.12 for those without dyslipidemia and 2.53 for those with dyslipidemia. The adjusted hazard ratio (95% confidence interval) for dyslipidemia was 0.856 (0.484-1.516) for men and 1.260 (0.661-2.400) for women; neither was significant when adjustment was made for age, systolic blood pressure, diastolic blood pressure, body mass index, creatinine clearance, diabetes mellitus, and proteinuria. CONCLUSIONS: The present study showed dyslipidemia to be an insignificant predictor of development of ESRD in the general Okinawa population.

Outcome study of renal biopsy patients in Okinawa, Japan.

Here we report a community-based epidemiologic study of patients who received renal biopsy in Okinawa, Japan between 1967 and 1994. The total number of cases was 2832 (1395 men and 1437 women), and the mean (SD) age at biopsy was 30.0 (10.0) years (range 1.0 to 88.0 years). The most common clinical indications for renal biopsy were proteinuria/hematuria (46.7%), nephrotic syndrome (21.2%), acute glomerulonephritis (10.1%), and systemic lupus erythematosus (7.5%). Patients who received renal biopsy between 1985 and 1994 (N= 1480) were much less likely to have acute glomerulonephritis than patients treated between 1967 and 1984 (N= 1352); the rates of proteinuria/hematuria, renal failure, and diabetes mellitus were slightly higher in the later period. Okinawa patients who began dialysis between 1971 and 2000 (N= 5246) were also studied. Among them, a total of 468 patients (260 men and 208 women) began dialysis after renal biopsy. The cumulative incidence of end-stage renal disease (ESRD) among these patients was 17% in 17 years. Half of these patients developed ESRD in the 5.8 years after renal biopsy. Among the dialysis patients, the biopsy rate was 12.6% in chronic glomerulonephritis, 1.7% in diabetes mellitus, 2.6% in nephrosclerosis, and 52.1% in systemic lupus erythematosus. The diagnoses of primary renal diseases were primarily made clinically. The survival rate after starting dialysis therapy was slightly better in those with than in those without renal biopsy but this finding was not statistically significant (adjusted hazards ratio 0.855, 95% CI 0.711-1.028, P= 0.095). The clinical significance of renal biopsy, other than its provision of histologic evidence, remains to be shown.

Determinants of prescribed dialysis dose and survival in a cohort of chronic hemodialysis patients.

BACKGROUND: The Determinants of the prescribed dialysis dose have not been well studied in a large patient population. Few studies have examined survival rates after adjusting for dose determinants. METHODS: Data were obtained from a cohort of chronic hemodialysis patients for the period January 1991 through December 2000. The prescribed dialysis dose was calculated as the dialyzer membrane area (m2) times session hours, and was expressed as m2h per week. Determinants of the prescribed dialysis dose were examined by multivariate logistic regression analysis of baseline clinical and laboratory variables. Survival curves for each prescribed dose were calculated by the Kaplan-Meier method. Cox proportional hazards analysis was used to evaluate differences in the survival curves after adjusting for confounding variables. The delivered dose of dialysis, Kt/V, was calculated in a subgroup of the cohort. RESULTS: For 1041 patients receiving thrice-weekly dialysis, the mean (SD) dialysis dose was 19.8 (5.8) m2h/week (range, 6.3 to 33.0 m2h/week). The significant and independent determinants of prescribed dialysis dose were sex, age, diabetes mellitus (DM), body mass index (BMI), serum albumin, diastolic blood pressure, serum creatinine, duration of dialysis, and comorbidity. The dialysis dose received by women and patients with DM was relatively low, even when adjusted for BMI ( P < 0.01 for both). During the follow-up period, 463 patients died, 60 underwent renal transplant, and 10 were transferred away from Okinawa. The hazard ratio (95% confidence interval) for death was 1.016 (0.995-1.037; not significant) for the dialysis dose (m2h/week) after adjustment for multiple confounding factors. The mean (SD) Kt/V was 1.31 (0.28). The hazard ratio (95% confidence interval) for Kt/V > or = 1.31 vs Kt/V < or = 1.30 was 0.706 (0.553-0.900; P = 0.0049). CONCLUSIONS: The prescribed dialysis dose did not significantly influence mortality in our cohort. Empirically based prescription practice, such as increasing the prescribed dialysis dose in male patients, when the BMI is large, or when serum creatinine or diastolic blood pressure is high may explain the relatively good prognosis of chronic hemodialysis patients in Japan.

Blood pressure predicts risk of developing end-stage renal disease in men and women.

Blood pressure as a risk factor for development of end-stage renal disease has not been fully studied, particularly in women. We studied the development of end-stage renal disease from 1983 through 2000 in 98 759 subjects, 46 881 men and 51 878 women, 20 to 98 years of age, who were screened in 1983 in Okinawa, Japan. Data for all dialysis patients registered from 1983 to 2000 in Okinawa were used to identify the screened subjects in whom end-stage renal disease developed. In follow-up, 400 subjects, 231 men and 169 women, had end-stage renal disease. Age, body mass index, and adjusted relative risk for systolic and diastolic blood pressure for both men and women were measured. When these results were compared with an optimal blood pressure, the relative risk of development of end-stage renal disease for those with high-normal blood pressure and hypertension were significant in both men and women. Hypertension is a significant risk factor for development of end-stage renal disease not only in men but also in women. Control of blood pressure within normal levels should be stressed as a strategy to prevent end-stage renal disease in both men and women.

Effect of the duration of dialysis on survival in a cohort of chronic haemodialysis patients.

BACKGROUND: Atherosclerosis and vascular calcification are common in chronic haemodialysis (HD) patients, and usually progress with time. Whether the length of dialysis treatment in chronic HD patients is a significant independent risk factor of death is not clear. METHODS: A cohort of chronic HD patients from the Okinawa Dialysis Study, n=1243 (720 men, 523 women), was followed from January 1991 to December 2000, and their survival rates were compared against the duration of HD, which was calculated in months from the start of dialysis therapy to January 1991. A Cox proportional hazards regression analysis was done to examine the influence of the duration of dialysis on survival, after adjusting for other factors such as age, sex, serum albumin concentration and diastolic blood pressure. The hazards ratio and 95% confidence interval (CI) were calculated in both diabetic and non-diabetic patients. RESULTS: The mean duration of dialysis was 61.9 months and ranged from 1 to 233 months. The numbers of patients who died, underwent renal transplantation or were transferred outside Okinawa were 568 (45.7%), 61 (4.9%) and 14 (1.1%), respectively, during the study. The hazards ratio (95% CI) was 1.002 (1.000-1.004, P=0.0245) for non-diabetic patients and 1.006 (1.001-1.011, P=0.0214) for diabetic patients, suggesting that the longer the duration of dialysis, the greater the risk of death. CONCLUSIONS: This study shows that prolonged dialysis is a significant predictor of death in chronic HD patients, in particular diabetic patients. Whether this is related to the progression of the atherosclerotic process or to uraemic conditions remains to be shown.

Impact of multiple risk factor clustering on the elevation of blood pressure.

A family history of hypertension, obesity, diabetes mellitus, hypercholesterolemia, and hypertriglyceridemia have all been associated with risk of hypertension. We retrospectively conducted a longitudinal study in a large screened cohort to explore the effect of the clustering of these five risk factors on the elevation of blood pressure (BP) in normotensive subjects at baseline. The study group comprised 4,857 normotensive subjects not treated with antihypertensive drugs (systolic BP < 140 mmHg, diastolic BP < 90 mmHg, 3,111 men and 1,746 women) who were followed up from 1997 to 1999. By 1999, 360 subjects had BP at the hypertensive level (systolic BP > or = 140 mmHg or diastolic BP > or = 90 mmHg). The incidence of subjects whose BP became hypertensive was 37 per 1,000 person-years. After adjusting for age, sex, systolic BP and other clinical factors, multiple logistic analysis showed that the relative risk of BP elevation was 1.49 (95% Cl: 1.09 to 2.05) in subjects with one risk factor; 1.65 (95% Cl: 1.15 to 2.27) in those with two risk factors; 1.42 (95% Cl: 0.91 to 2.32) in those with three; and 4.86 (95% Cl: 2.58 to 9.16) in those with four or more when compared with subjects with no risk factors. Multiple regression analysis showed that the number of risk factors was positively correlated with an increase in BP from 1997 to 1999; the regression coefficient was 0.51 (p = 0.001) for increase in systolic BP, and 0.31 (p = 0.008) for increase in diastolic BP after adjusting for age and sex. In conclusion, clustering of risk factors significantly predicted the development of hypertension.

Evidence for elevated pulse pressure in patients on chronic hemodialysis: a case-control study.

BACKGROUND: Few analyses have compared pulse pressure (PP) values in hemodialysis patients with healthy individuals, and they have provided only limited data. We retrospectively examined PP in a large cohort of hemodialysis patients and healthy control subjects. METHODS: The relationships of systolic blood pressure (SBP), diastolic blood pressure (DBP), and PP to mean arterial pressure (MAP) levels were investigated in 234 chronic hemodialysis patients and in 682 control subjects matched for age, sex, diabetes mellitus, and body mass index. RESULTS: In both control and patients, PP was positively correlated with MAP, and the two regression lines were parallel (beta of control subjects = 0.52; beta of hemodialysis patients = 0.57, P = 0.48). According to the regression line, at any MAP level, the PP in hemodialysis patients was significantly higher than that in control subjects: the mean PP difference between control and patients was 19.2 mm Hg (95% CI, 17.2 to 21.1 mm Hg, P < 0.0001). When the relationships between MAP and SBP and that between MAP and DBP were analyzed, the regression lines were also parallel. However, at any MAP level, SBP was higher and DBP was lower in hemodialysis patients than control subjects; the mean SBP difference was 12.8 mm Hg (95% CI, 11.5 to 14.1 mm Hg, P < 0.0001) and mean DBP difference was 6.4 mm Hg (95% CI, 5.7 to 7.0 mm Hg, P < 0.0001). CONCLUSIONS: At any MAP level, hemodialysis patients had a higher SBP, lower DBP, and higher PP values than those control subjects with a normal renal function who were matched for age, sex, diabetes mellitus, and body mass index. Further study is needed to determine whether preventing or reducing an elevated PP improves the prognosis for hemodialysis patients.

Severe abdominal pain associated with allergic reaction to nafamostat mesilate in a chronic hemodialysis patient.

A 33-year-old woman was referred from an outside dialysis clinic to our hospital because of severe abdominal pain during hemodialysis. She had been on chronic hemodialysis for the past 11 years due to chronic glomerulonephritis. Nafamostat mesilate was used as an anticoagulant for hemodialysis, because it was during her menstrual period with hypermenorrhea. On admission, she had no abdominal pain or gynecological abnormalities. On the second day, she had similar abdominal pain during hemodialysis with nafamostat mesilate in our dialysis unit. The abdominal pain disappeared within 60 minutes after discontinuing the hemodialysis. We re-started dialysis using heparin instead of nafamostat mesilate and she had no symptoms. The titer of total immunoglobulin E was high. The drug lymphocyte stimulation test was positive for nafamostat mesilate and antigen specific immunoglobulin E to nafamostat mesilate was highly positive in her blood. Although an allergic reaction to nafamostat mesilate is a rare complication, it should be one of the differential diagnoses of abdominal pain occurring during hemodialysis.

Triglyceride, but not total cholesterol or low-density lipoprotein cholesterol levels, predict development of proteinuria.

BACKGROUND: Epidemiological data about the relationship between dyslipidemia and proteinuria are sparse. We conducted a retrospective and longitudinal study in a large screened cohort to evaluate whether triglyceride, high-density lipoprotein (HDL) cholesterol, total cholesterol, and low-density lipoprotein (LDL) cholesterol levels increase the risk of development of proteinuria and loss of renal function. METHODS: Post hoc analysis was performed for 4326 subjects who were free from proteinuria (dipstick 1+ or higher) at baseline (1997) with a follow-up period through 2000. Outcome measures were the development of proteinuria (1+ or higher) and change in glomerular filtration rate (GFR). Multiple logistic analysis and multiple regression analysis were used to analyze baseline characteristics related to the outcome measures. RESULTS: During the observational period, 505 (11.7%) of subjects had one or more episodes of proteinuria (>/=1+). Adjusted relative risk of triglycerides for one or more incidences of proteinuria was 1.007 (95% CI 1.000 to 1.105, P = 0.04) in men and 1.032 (95% CI 1.004 to 1.061, P = 0.02) in women. Total cholesterol, HDL cholesterol, and LDL cholesterol were not significant predictors of proteinuria. The mean change in GFR between 1997 and 2000 was -6.3 (SD = 9.0) mL/min/1.73 m2 in men, and -7.8 (SD = 10.7) mL/min/1.73 m2 in women. HDL cholesterol (beta = 0.04, t = 3.7, P = 0.0002) in men and triglycerides (per 10 mg/dL, beta = -0.09, t = -2.2, P = 0.02) in women were correlated with the change in GFR. CONCLUSIONS: High triglyceride levels predicted a risk of developing proteinuria in both men and women, but not total cholesterol nor LDL cholesterol. High triglyceride in women and low HDL cholesterol in men predicted the decline of renal function. It remains to be determined whether prospective treatment of dyslipidemia will protect against renal injury.

Analysis of drug prescription in chronic haemodialysis patients.

BACKGROUND: Information concerning medication use in Asian haemodialysis patients is sparse. We surveyed prescribed medications and examined the relation between the number of medications and mortality and clinical characteristics in chronic haemodialysis patients, in Okinawa, Japan. METHODS: We conducted a cross-sectional multicentre survey in August 1999 and patients were observed during 13 months of follow up. RESULTS: The clinical demographics of 850 chronic haemodialysis patients in seven dialysis units were obtained. Compared with the mean number of medications prescribed in ambulatory patients treated in general practice reported from Ministry of Health and Welfare of Japan (2.7 (n=20 716)), the mean number medications in haemodialysis patients was larger (7.2 (n=850)). The three most prescribed drug types in haemodialysis patients were those related to calcium and phosphate metabolism (88%), antihypertensive agents (71%), and erythropoietin (60%). Among the 850 patients, 38 died during the 13-month follow-up period. The number of medications was positively associated with mortality after adjusting for age, sex, and other clinical factors: the hazard ratio was 1.14 (95% confidence interval 1.03-1.26, P=0.007). A multiple linear regression analysis using the number of medications as a dependent factor and sex and other clinical characteristics as independent factors revealed that male sex (P=0.04), diabetes mellitus (P<0.0001), and duplication of drugs (P<0.0001) were positively correlated with the number of medications. CONCLUSIONS: Multiple drug use was observed in haemodialysis patients. The number of prescribed drugs was a significant predictor of short-term mortality. Male sex, diabetes mellitus, and duplication of drugs were correlated with increases in the number of medications.

Influence of smoking and obesity on the development of proteinuria.

BACKGROUND: Proteinuria is a significant risk factor for end-stage renal disease. Previous evidence suggested that smoking and obesity increase the risk of proteinuria. However, it is unclear whether these risk factors predict the development of proteinuria independently of hypertension and diabetes mellitus. The aim of this study was to analyze the effects of obesity and smoking on the development of proteinuria in a screened cohort of subjects with normal kidney function. METHODS: A total of 5403 subjects (3403 men and 2000 women) who participated in the 1997 and 1999 health screening examinations in Okinawa Japan, and who were normal renal function (serum creatinine < or =1.2 mg/dL in men, < or =1.0 mg/dL in women) and negative proteinuria by dipstick examination in 1997 were eligible for study. Logistic analysis was used to examine the relation between the baseline state of smoking or obesity in 1997, and the development of proteinuria in 1999, adjusted for age, sex, and other confounding factors. RESULTS: Proteinuria developed in 5.8% of participants (6.7% in men, 4.4% in women; dipstick score, 1+ in 277, 2+ in 37, and > or =3+ in 4 participants). The incidence of proteinuria was positively associated with the number of cigarettes smoked per day (P = 0.04), and a body mass index (P < 0.0001) at baseline. Analysis showed that the relative risk (95% confidence interval) of developing proteinuria was 1.32 (1.00 to 1.74), P = 0.04 for cigarette smoking, 1.45 (1.13 to 1.86), P = 0.002 for obesity, 1.56 (1.19 to 2.06), P = 0.001 for hypertension, and 2.27 (1.55 to 3.32), P < 0.0001 for diabetes mellitus. Stratified with men and women, the relative risk was 1.28 (0.96 to 1.72), P = 0.09 for smoking, and 1.60 (1.19 to 2.14), P = 0.001 for obesity in men; the relative risk was 1.30 (0.44 to 3.80), P = 0.62 for smoking, and 1.04 (0.63 to 1.72), P = 0.87 for obesity in women. CONCLUSIONS: Hypertension and diabetes mellitus were superior to smoking and obesity in predicting the development of proteinuria in all subjects. Stratified with men and women, obesity was a significant risk factor for the development of proteinuria independently of both hypertension and diabetes mellitus in men. The risk of developing proteinuria also tended to be increased with cigarette smoking in men. Smoking and obesity in women were not significant in this data set.

Prevalence and correlates of diabetes mellitus in a screened cohort in Okinawa, Japan.

The incidence of end-stage renal disease due to diabetes mellitus (DM) is increasing. There have been too few epidemiological studies of the predictors of DM nephropathy, particularly type 2 DM, among a statistically significant population. We studied the prevalence and correlates of DM in a screened cohort in Okinawa, Japan. A total of 9,914 screenees (6,163 men and 3,751 women) over 18 years of age underwent a 1-day health check at the Okinawa General Health Maintenance Association between April 1997 and March 1998. Subjects were considered to have DM if they showed a fasting plasma glucose > or = 126 mg/dl and hemoglobin A1c > or = 7.0%, or if they were receiving treatment for DM. Non-DM subjects were followed-up until March 2000 to see whether or not they developed DM. Relative risk for developing DM was evaluated by Cox proportional hazard analysis after adjusting for confounding variables. A total of 673 screenees (520 men and 153 women) were diagnosed with DM. The prevalence of DM was 67.9 per 1,000 screenees (84.4 for men and 40.8 for women). A total of 7,125 non-DM screenees were examined a second time. Among them, 164 screenees (130 men and 34 women) had developed DM during the follow-up period. Over 2 years, the cumulative incidence of DM was 2.3% (2.9% in men and 1.3% in women). The adjusted relative risk (95% confidence interval) for developing DM was highest for proteinuria, or 1.90 (1.14-3.17). The results indicated that the prevalence and incidence of DM were high among this screened cohort in Okinawa, Japan. Subjects with proteinuria may thus be at high risk for developing DM.

Hypocholesterolemia is a significant predictor of death in a cohort of chronic hemodialysis patients.

BACKGROUND: Although hypocholesterolemia is common in chronic hemodialysis patients, its effect on survival has not been studied in a large patients population. METHODS: A cohort of chronic hemodialysis patients (N = 1167) was prospectively followed from January 1991 to January 2001. The survival impact of this cohort, who were divided according to different baseline levels of serum cholesterol, were calculated with the multivariate Cox proportional hazard analysis after adjusting for baseline clinical and laboratory variables. RESULTS: During the study period, 567 (48.6%) patients died. The mean (SD) baseline level of serum cholesterol was 171.0 (40.8) mg/dL and ranged from 76 to 378 mg/dL. The five-year survival rate was highest (0.812) in the subgroup that had a serum cholesterol range of 200 to 219 mg/dL and was lowest (0.608) in the subgroup with serum cholesterol values of <140 mg/dL. The five-year survival rate was 0.735 in the subgroup with serum cholesterol of > or =220 mg/dL. Serum cholesterol was a significant predictor of death with an adjusted hazards ratio (95% confidence interval) was 0.939 (0.891 to 0.989). In a subgroup of patients with serum albumin values > or =4.5 g/dL (N = 128), the adjusted hazards ratio was even greater at 1.370 (1.105 to 1.692). Other than sex, body mass index and serum albumin were significant determinants of baseline levels of serum cholesterol. CONCLUSIONS: Hypocholesterolemia was an independent predictor of death in patients on chronic hemodialysis. This impact of hypercholesterolemia on survival was only evident in a subgroup of patients whose serum albumin was more than 4.5 g/dL.

End-stage renal disease associated with familial Mediterranean fever.

A 39-year-old man had been suffering from periodic fever since childhood. He was started on hemodialysis due to secondary amyloidosis on December 2000. The patient was believed to have Familial Mediterranean fever (FMF) because of recurrent fever with peritonitis, arthritis and inflammatory changes and secondary amyloidosis in his kidneys, heart and colon. No other family member had recurrent fever. IL-6, TNF, and dopamine beta-hydroxylase were not increased in the febril phase. The patient was homozygous for the M6941 mutation. We report the first Japanese case of FMF associated with amyloidosis and confirmed by a gene mutation.

Demographic trends in the Okinawa Dialysis Study (OKIDS) registry (1971-2000).

BACKGROUND: The clinical demographics of chronic dialysis patients are changing worldwide. However, long-term data from regional dialysis registries have not yet been analyzed and reported. METHODS: The Okinawa Dialysis Study (OKIDS) registry included all chronic dialysis patients treated in Okinawa, Japan, since 1971. Data for the years 1971 to 1990 were analyzed to predict trends for 1991 to 2000. The predicted values were then compared to the actual values and analyzed statistically, with particular attention being paid to relative risk of death. Multivariate Cox proportional hazards analysis was done to analyze the time factors of relative risk of death. RESULTS: A total of 5246 patients (2981 men and 2265 women) were registered and the total duration of observation was 28,431 patient-years. The prevalence and incidence of dialysis patients expressed per million population were 2320 and 297, respectively, in 2000, values that were significantly higher (P < 0.02 for both) than the predicted values. The gross mortality rate per 1000 patient-years was 118.4 for 1971 to 1980, 63.3 for 1981 to 1990, and 77.7 for 1991 to 2000. The adjusted hazards ratio (95% confidence interval) for mortality was 0.743 (0.650 to 0.862) for 1981-1990 and 0.721 (0.659 to 0.790) for 1991 to 2000 in comparison to the risk of mortality in 1971 to 1980. The decrease in mortality rate was largely due to the drop in cardiac deaths from 71.0 for 1971 to 1980 to 17.2 for 1991 to 2000. CONCLUSIONS: The incidence and prevalence of chronic dialysis patients increased more than expected over the past decade in Okinawa, Japan. Despite the rapid change in patient demographics, the survival rate did not decrease significantly.

Pulse pressure and risk of total mortality and cardiovascular events in patients on chronic hemodialysis.

BACKGROUND: Pulse pressure (PP) has been shown as a risk factor for mortality or cardiovascular events in several studies. However, the impact of PP on prognosis in a cohort of chronic hemodialysis patients has not been sufficiently studied. We examined the effect of PP on total mortality and cardiovascular events in chronic hemodialysis patients, and whether PP adds useful value to systolic blood pressure (SBP) or diastolic blood pressure (DBP) for predicting total mortality and cardiovascular events in chronic hemodialysis patients. METHODS: Chronic hemodialysis patients (N=1243, 720 men, 523 women) alive on January 1, 1991 at baseline were involved in this study. Cox regression, adjusted for age, sex, and other risk factors, was used to assess the relation between blood pressure components and risk of death and cardiovascular events over a nine-year follow-up. RESULTS: The association with the risk of total mortality was positive for PP (P=0.002) and SBP (P=0.04), but not significant for DBP (P=0.4), considering each pressure individually (single blood pressure component model, SPM); of the three measurements, PP yielded the highest chi2 value. When SBP and DBP were jointly entered into the Cox regression model (dual blood pressure component model, DPM), the association with the risk of total mortality was positive for SBP (HR, 1.083; 95% CI, 1.030 to 1.137) and negative for DBP (HR, 0.886; 0.808 to 0.970). After the addition of diabetes mellitus as an adjusted variable to the model, PP was not a significant predictor for total mortality; PP was a significant predictor for total mortality in non-diabetic patients, but not in diabetic patients. PP was positively associated with the risk of stroke, and stroke and AMI; however, predictive value of PP for each endpoint was not superior to SBP and DBP in SPM. In DPM with SBP and DBP, the association with the risk of stroke and acute myocardial infarction (AMI) was positive for SBP (P=0.02) but not significant for DBP (P=0.5). In DPM with SBP and PP, the association with the risk of stroke and AMI was positive for SBP (P=0.01) but not significant for PP (P=0.5). CONCLUSIONS: In non-diabetic patients on chronic hemodialysis, PP was an independent predictor of total mortality. PP was more potent predictor of total mortality than SBP or DBP. For predicting cardiovascular events, SBP was superior to PP or DBP.