There may be cumulative 'metabolic scars' after a decade or more of long-term hemodialysis. We studied 39 patients who have been on maintenance hemodialysis for 10-24 years to determine their functional status and pathobiology. The 39 long-term (greater than or equal to 10 years) patients were compared with a control cohort of 37 age, gender-, race-, and renal-diagnosis-matched patients on hemodialysis for less than or equal to 3 years. The functional status was measured using a modified Karnofsky scale, and the employment status was noted as well. Details of hospitalizations and intercurrent infections requiring outpatient oral or intravenous antibiotic therapy during the preceding year were obtained. Comorbid medical conditions were documented, and basic laboratory tests were performed. The mean age of the long-term patients was 51.8 +/- (SE) 1.9 years, and the mean age of the control group was 51.5 +/- 2.4 years (p = 0.92). Three times weekly hemodialysis prescriptions were similar in both groups (long-term: 3.5 +/- 0.02 h, control: 3.4 +/- 0.02 h; p = 0.27). The mean modified Karnofsky scores were equivalent in both groups. Also the rates of employment were poor and equivalent in both groups. The rate of hospitalizations during the preceding year was higher among the long-term patients (0.92 +/- 0.19/patient year) than in the control patients (0.51 +/- 0.15/patient year; p = 0.09). The long-term patients had more intercurrent infections (1.23 +/- 0.21) than the controls (0.68 +/- 0.16; p = 0.04). Comorbid medical conditions were more prevalent in the long-term patients as reflected by their comorbidity index of 1.6 +/- 0.2 versus a comorbidity index of 0.7 +/- 0.16 for the control patients (p = 0.007). Mean postdialysis weight and serum albumin and serum creatinine concentrations were equivalent in both groups. Of all the measured laboratory indices, only the hematocrit differed significantly between the two groups (long-term patients: 29 +/- 0.79%, controls: 26.6 +/- 0.69%; p = 0.02), although both groups received equivalent three times weekly doses of recombinant human erythropoietin. We conclude that patients on maintenance hemodialysis for 10 years or more have more intercurrent infections as well as more comorbid medical conditions than freshly started (1-3 years) hemodialysis patients. Also there is a clinically significant increased risk of hospitalization among the long-term patients. However, nutritional status, level of physical activity, and rate of employment in the long-term patients were equivalent to that of freshly started hemodialysis patients.