|
|
| |||||||||||||||||||||||||||||||
Background Alcohol abuse may be accompanied by a variety of disorders of electrolyte and acid-base metabolism. The role of the kidney in the pathogenesis of these disturbances is obscure. We sought to evaluate the alcohol-induced abnormalities of renal function and improvement during abstinence and to assess the relation between renal dysfunction and electrolyte and acid-base disorders.
Methods We measured biochemical constituents of blood and renal function before and after four weeks of abstinence in 61 patients with chronic alcoholism who had little or no liver disease.
Results On admission, 18 patients (30 percent) had hypophosphatemia and hypomagnesemia, 13 patients (21 percent) had hypocalcemia, and 8 patients (13 percent) had hypokalemia. Twenty-two patients (36 percent) had a variety of simple and mixed acid-base disorders. Twenty of these patients had metabolic acidosis, and among them, 80 percent had alcoholic acidosis. A wide range of defects in renal tubular function, with normal glomerular filtration rate, was detected in these patients. The defects included decreases in the threshold and maximal reabsorptive ability for glucose (38 percent of patients) and in the renal threshold for phosphate excretion (36 percent); increases in the fractional excretion of
Conclusions Transient defects in renal tubular function are common in patients with chronic alcoholism and may contribute to their abnormalities of serum electrolyte and blood acid-base profiles.
2-microglobulin (38 percent), uric acid (12 percent), calcium (23 percent), and magnesium (21 percent); and aminoaciduria (38 percent). Seventeen patients (28 percent) had a defect in tubular acidification, and five an impairment in urinary concentrating ability. Urinary excretion of N-acetyl--d-glucosaminidase and alanine aminopeptidase was increased in 41 and 34 percent of patients, respectively. The abnormalities of blood chemistry and renal tubular function disappeared after four weeks of abstinence.
Methods
Patients
We studied 61 patients with chronic alcoholism who volunteered for (and completed) four weeks of withdrawal therapy. To be eligible, the patients had to have had a large intake of alcohol for at least five years, a weekly alcohol consumption of 600 g or more for the previous three months, no histopathological evidence of cirrhosis or alcoholic hepatitis, no clinical or laboratory evidence of pancreatitis or malnutrition, no history of renal disease or exposure to lead or cadmium (from either occupational exposure or consumption of illicitly distilled spirits), and no laboratory evidence of excessive body stores of lead (calcium-sodium-EDTA lead-mobilization test) or cadmium. We excluded patients admitted for alcohol-related diseases, such as alcoholic liver disease, pancreatitis, gastrointestinal bleeding, acute atrial fibrillation, cardiomyopathy, seizures, and rhabdomyolysis, as well as those with vomiting, diarrhea, or intercurrent illness that could influence acid-base and electrolyte homeostasis. Patients in whom severe symptoms of alcohol withdrawal developed were also excluded, as were those who required medications, such as phenobarbital, antihypertensive drugs, or fluid-replacement therapy. None of the patients had delirium tremens or alcoholic hallucinosis, and none were receiving medications known to influence renal function or mineral and electrolyte metabolism. Each patient had been drinking until 24 hours before admission, although none were intoxicated at the time of entry into the study. We also studied 42 normal subjects who abstained from alcohol or consumed only small amounts (15 to 105 g of ethanol per week). All the subjects gave informed written consent for the study.
Study Design
The patients were allowed to move about freely in the hospital but not to leave it during the four-week study period. Renal function, concentrations of serum electrolytes, and blood acid-base values were determined immediately after admission and after four weeks of abstinence. Each patient's nutritional state was evaluated on admission, and the patients were asked about their alcohol consumption. Renal tubular function was estimated on the basis of a four-day protocol. Arterial blood gases and serum electrolyte concentrations were also measured on the fourth and seventh days of hospitalization, and the fractional urinary excretion of electrolytes was determined at the same times. A percutaneous liver biopsy was performed within two weeks after admission; the microscopical sections were evaluated by a pathologist unaware of the patient's clinical status. During the study period, the patients were fed a standard diet that provided 140 mmol of sodium daily. Protein intake was not restricted. All the patients received B vitamins and folic acid, but none received electrolyte supplements.
Assessment of Alcohol Consumption and Nutritional State
Alcohol consumption was assessed with a structured questionnaire, and the results were corroborated by an in-depth interview performed by a trained interviewer. The total weekly alcohol consumption was calculated and expressed as grams of ethanol. In addition, biologic markers of alcohol intake, such as mean corpuscular volume and serum concentrations of gamma-glutamyltransferase, aspartate aminotransferase, and alanine aminotransferase were measured in all patients5. Nutritional state was estimated by anthropometric measurements and laboratory indexes (hemoglobin, blood lymphocytes, and serum concentrations of total protein, albumin, urea, and creatinine). Skin-fold thickness was measured at four sites with standard calipers, and estimates of body fat were obtained with appropriate equations based on age and sex.
Assessment of Tubular Function
Renal function was estimated on the basis of a four-day protocol conducted immediately after admission and after four weeks of abstinence. On the first day we measured the glomerular filtration rate with tests of inulin clearance and endogenous creatinine clearance; the fractional urinary excretion of sodium, potassium, chloride, calcium, and magnesium; and the theoretical renal threshold for phosphate excretion. Fasting urine pH and osmolality were also measured. A 24-hour urine specimen was collected for the determination of proteins, glucose, ammonium, titratable acidity, bicarbonate, and amino acids, as well as electrolytes, creatinine, N-acetyl--d-glucosaminidase, 
-glucosidase, and alanine aminopeptidase. On the second day, each patient was given 4 g of sodium bicarbonate with 250 ml of water, after which urine was collected for two hours and a blood sample was collected one hour after the administration of sodium bicarbonate for 2-microglobulin assay. Urine-concentrating ability was assessed on the third day by measurements of plasma and urinary osmolality during water deprivation. After 12 hours of water deprivation, 5 units of vasopressin (desmopressin acetate [DDAVP, USV Pharmaceuticals]) was given subcutaneously, after which hourly urine samples were collected for 4 hours for measurement of urinary osmolality. A maximal urinary osmolality of 
750 mOsm per kilogram after vasopressin administration was considered a normal response6. Patients whose fasting urinary osmolality exceeded 750 mOsm per kilogram were considered to have normal concentrating ability and did not undergo the water-deprivation and vasopressin tests. The acidification ability of the distal renal tubules was tested by a short acid-loading test on the fourth day. Calcium chloride (2 mmol per kilogram of body weight) was administered orally over a 60-minute period, and urine and blood pH was measured every 2 hours for 6 hours7. If the fasting urine pH was less than 5.5, regardless of the blood pH, the patient was considered to have normal acidification ability, and the acid-loading test was not performed. Finally, a glucose-titration study was performed8 to determine the threshold and the maximal reabsorptive ability for glucose.
Assessment of Plasma Hormone Concentrations
After the patient had been recumbent and fasting overnight, blood samples were collected on the first morning after admission for the estimation of plasma concentrations of aldosterone, parathyroid hormone, cortisol, corticotropin, norepinephrine, and epinephrine and of plasma renin activity. Further blood samples were obtained for plasma cortisol measurements at 4 p.m., and midnight that day. The hormone study was repeated after the patients had been abstinent for four weeks.
Laboratory Measurements
Routine biochemical determinations were performed in serum and urine by standard automated methods. Inulin was determined photometrically with indolacetic acid after hydrolyzation to fructose. The unmeasured-anion concentration was calculated as the sodium concentration minus the sum of the chloride and bicarbonate concentrations. Standard formulas were used to calculate the fractional urinary excretion of substances. Plasma and urinary osmolality was determined by freezing-point depression. Arterial-blood pH and partial pressure of carbon dioxide (PaCO2) and urine pH were measured with a digital acid-base analyzer. Urinary ammonium, titratable acidity, and total carbon dioxide content were measured by standard techniques. Serum and urine 2-microglobulin was measured with Phadebas 2 microtest kits (Pharmacia Diagnostics). Urinary excretion of N-acetyl--d-glucosaminidase, alanine aminopeptidase, and -glucosidase was determined by colorimetric methods (Far Diagnostici). Urinary amino acids were measured by high-performance liquid chromatography. Serum -hydroxybutyrate was assayed with a procedure based on the method of Williamson et al9. Serum acetoacetate was measured semiquantitatively by the nitroprusside reaction, and urine ketones by dipstick. Lead and cadmium were measured in plasma and urine by electrothermal atomic-absorption spectrometry. Plasma concentrations of norepinephrine and epinephrine were determined by a modification of the radioenzymatic method of Da Prada and Zurcher10. Plasma concentrations of aldosterone, cortisol, and corticotropin and plasma renin activity were measured by radioimmunoassay. Plasma parathyroid hormone was measured with a two-site immunoradiometric assay.
Statistical Analysis
The results are expressed as means ±SD. Comparisons within and between groups were analyzed by analysis of variance, the Wilcoxon rank-sum test, and Student's t-test, as appropriate. Correlation was assessed with simple and multiple linear regression and Spearman's rank test. P values less than 0.05 were considered to indicate statistical significance.
Results
Drinking Habits, Laboratory Markers of Alcoholism, and Liver Damage
Some characteristics of the patients with chronic alcoholism and the normal subjects are shown in Table 1. The values for the biologic markers of alcohol intake in the alcoholic patients are shown in Table 2. The liver-biopsy specimens were normal or showed only minor, nonspecific changes in 45 patients. They showed moderate steatosis in 6 patients, with and without fibrosis (3 patients each), and showed mildly fatty liver in 10 patients. The values for the biologic markers were similar in the 45 patients with no or minor histologic abnormalities and in all 61 patients.
|
|
In the alcoholic patients the mean values for body-mass index and body fat were slightly, but not significantly, higher than those in the normal subjects (Table 1). Conversely, there was no significant difference in any of the laboratory indexes of nutritional status (Table 2).
Electrolyte and Acid-Base Disorders
Several patients had symptoms and signs attributable to serum electrolyte or blood acid-base disturbances. Generalized muscular weakness was a common finding and could have resulted from hypophosphatemia, hypocalcemia, or hypomagnesemia. Tachypnea and hyperpnea were present in the patients with severe metabolic acidosis. The serum electrolyte and blood acid-base values are shown in Table 3. As compared with the normal subjects, the alcoholic patients had lower serum concentrations of phosphate, potassium, magnesium, and ionized calcium on admission, lower values for blood pH and PaCO2, and lower plasma bicarbonate concentrations, whereas there was no significant difference between the groups in serum sodium and chloride concentrations. Among the alcoholic patients, 18 (30 percent) had hypophosphatemia, 18 (30 percent) had hypomagnesemia, 13 (21 percent) had hypocalcemia, and 8 (13 percent) had hypokalemia. The serum uric acid concentrations of these patients were similar to those of the normal subjects. Six patients (10 percent) had hyperuricemia (serum uric acid concentration, >7.0 mg per deciliter [420 µmol per liter] in men and 6.2 mg per deciliter [370 µmol per liter] in women), and seven patients (11 percent) had decreased serum uric acid concentrations (<3.2 mg per deciliter [190 µmol per liter]).
|
-hydroxybutyrate concentrations (3540 ±270 µmol per liter; normal, <100). The nitroprusside reaction was positive in 13 of these patients, and 12 patients had ketone bodies in their urine. Renal Tubular Dysfunction
The mean glomerular filtration rate in the alcoholic patients, as estimated from the inulin clearance, did not differ from that of the normal subjects (Table 4). The mean values for creatinine clearance were also similar. None of the patients had a 24-hour urinary protein excretion exceeding 0.15 g. The alcoholic patients had a variety of transport abnormalities of renal tubules on admission (Table 4). Their mean fractional excretion of 2-microglobulin was higher than that of the normal subjects, and in 23 of the 61 alcoholic patients it exceeded the mean value in the normal subjects by more than 2 SD. The patients' serum concentrations of 2-microglobulin ranged from 0.9 to 2.6 mg per liter (mean, 1.8 ±0.5), as compared with 2.2 ±0.5 mg per liter in the normal subjects (P = 0.004). The mean values for the excretion threshold and the maximal reabsorptive ability for glucose were lower than those of the normal subjects, and 23 patients had values for these measures that were more than 2 SD below the mean in the normal subjects. Twelve of these 23 patients had glycosuria, with glucose levels ranging from 1.5 to 11.5 g in a 24-hour urine sample. All the patients with renal glycosuria had increased urinary excretion of amino acids. The pattern of excretion was generalized, including neutral, acidic, and basic amino acids.
|
|
In the alcoholic patients, the urinary excretion of N-acetyl--d-glucosaminidase, alanine aminopeptidase, and -glucosidase was higher than that in the normal subjects (Table 4).
Relation between Tubular Dysfunction and Blood Chemistry
Tubular dysfunction appeared to contribute to the abnormalities of serum electrolyte and blood acid-base profiles in the alcoholic patients. There was a positive correlation (P<0.001) between the serum phosphate concentration and the renal threshold for phosphate excretion, and there were inverse correlations between the serum concentrations of ionized calcium (P = 0.002), magnesium (P = 0.008), and uric acid (P<0.001) and their fractional urinary excretion, measured simultaneously. The defects in tubular function were more common in patients with corresponding abnormalities of serum electrolytes. Specifically, 78 percent of the patients with hypophosphatemia had a low renal threshold for phosphate excretion, and 62 percent of the patients with hypocalcemia, 50 percent of those with hypomagnesemia, and 50 percent of those with hypokalemia had increased fractional urinary excretion of the respective substance. The percentages of patients with normal serum electrolyte concentrations who had these abnormalities were 9, 10, 9, and 8 percent, respectively. Six of seven patients with hypouricemia (86 percent) had increased fractional excretion of uric acid. Eleven patients with metabolic acidosis (55 percent) had impaired renal acidification ability and reduced net acid excretion. Six of 10 patients in whom metabolic acidosis was not accompanied by respiratory alkalosis, metabolic alkalosis, or both had abnormally low net acid excretion despite systemic acidemia. Finally, of the six patients with renal tubular acidosis, two (33 percent) had hypokalemia and three (50 percent) had hypophosphatemia.
Effects of Abstinence on Serum Electrolyte and Blood Acid-Base Disorders
During hospitalization and the period of abstinence the mean serum concentrations of potassium, magnesium, phosphate, and ionized calcium increased, being significantly higher after four and seven days and four weeks than on day 1 (Table 6). There was parallel improvement of acid-base abnormalities (Table 3). On day 7 all the patients with wide-anion-gap metabolic acidosis and coexisting respiratory or metabolic alkalosis had recovered, but three patients still had normal-anion-gap hyperchloremic acidosis.
|
The abnormalities of renal tubular function improved with abstinence, and after four weeks the frequency of abnormal findings was markedly decreased (Table 4). The fractional excretion of potassium decreased quickly, with significantly lower values on days 4 and 7 than on day 1 (Table 7); at these times, the values were clearly subnormal (P<0.001). There was a parallel increase in the renal threshold for phosphate excretion during the first days of hospitalization. The fractional excretion of calcium and magnesium decreased more slowly. After four weeks of abstinence, six patients had decreases in the threshold and the maximal reabsorptive ability for glucose. Urinalysis revealed generalized aminoaciduria in five of these patients. The disturbances in renal acidification disappeared almost completely (Table 5), and only three patients did not have normal urinary acidification after calcium chloride loading. The abnormal responses to water deprivation and vasopressin persisted in two patients.
|
On admission, the alcoholic patients had higher mean plasma renin activity and higher plasma concentrations of aldosterone, norepinephrine, epinephrine, and cortisol than did the normal subjects (Table 8). The mean plasma concentration of parathyroid hormone in the alcoholic patients was lower than that in the normal subjects, and all the patients with hypocalcemia or hypomagnesemia had low plasma concentrations of parathyroid hormone. After four weeks of abstinence, their values were similar to those in the normal subjects, except that the mean plasma concentration of parathyroid hormone was still lower than that in the normal subjects.
|
We found that patients with chronic alcoholism have a variety of renal tubular abnormalities that are independent of chronic liver disease, pancreatitis, and rhabdomyolysis and that occur in the presence of normal glomerular filtration. The abnormalities were reversible, in most instances disappearing after four weeks of abstinence, despite many years of alcohol abuse. The fact that alcoholism is associated with multiple abnormalities of renal tubules, involving different segments of the nephron, leads to the conclusion that exposure to ethanol may cause generalized tubular dysfunction. The presence of glycosuria, aminoaciduria, and a decreased renal threshold for phosphate excretion suggests that ethanol abuse may result in a generalized reduction in the reabsorptive ability of the proximal tubular cells. This hypothesis is supported by recent studies indicating that ethanol interferes with the carrier functions of these cells by decreasing Na+/K+-ATPase activity12,13,14. Thirty-eight percent of the patients with chronic alcoholism had increased fractional excretion of 2-microglobulin, another disturbance indicating abnormal proximal tubular function15. The increases in the urinary excretion of N-acetyl--d-glucosaminidase, a lysosomal enzyme from the proximal tubules, and of alanine aminopeptidase, a brush-border enzyme, also suggested the presence of proximal tubular defects16. Rat-kidney tissue contains alcohol dehydrogenase activity that appears to be similar or identical to that of the class I alcohol dehydrogenase in liver17,18,19. The oxidation of ethanol by alcohol dehydrogenase generates acetaldehyde, which may inhibit the activity of several enzymes, causing the cells to become less efficient14,20. In addition, the oxidation of acetaldehyde, especially by the acetaldehyde dehydrogenase, that has a lower Michaelis-Menten constant, generates species of free radical oxygen that are capable of damaging cell membranes21,22. In patients with chronic alcoholism these mechanisms may result in increased urinary excretion of enzymes and proximal tubular dysfunction.
Fractional calcium excretion was increased in one fourth of our patients, in whom serum concentrations of ionized calcium were slightly reduced. The effect of ethanol in decreasing the Na+/K+-ATPase activity in the proximal tubular cells12 may result in a decrease in the tubular reabsorption of calcium. The suppressed secretion of parathyroid hormone would further contribute to the decreased tubular reabsorption of calcium. The reduced secretion of parathyroid hormone may also have caused the enhanced fractional excretion of magnesium23 found in 21 percent of the alcoholic patients, to which the increased plasma concentrations of cortisol and aldosterone may have contributed24.
During the first days of hospitalization, the fractional excretion of potassium decreased to subnormal values, suggesting the possibility of a transport defect in the distal nephron. The fact that nearly 30 percent of the patients had an inability to acidify their urine indicates impairment in the function of the collecting duct. Our results are coincident with recent studies of the effects of prenatal exposure to ethanol on postnatal renal function. Rats exposed to ethanol during fetal life had defects in potassium excretion when given a potassium load25. Decreased fractional excretion of potassium, incomplete renal tubular acidosis, and impaired urine-concentrating ability are the main features of renal tubular dysfunction in infants with the fetal alcohol syndrome26. Although the molecular basis of the tubular acidification defect remains to be established, some pathophysiologic mechanisms may be hypothesized. The decrease in urinary excretion of titratable acids may be a consequence of the inappropriately elevated urine pH. The diminished rate of phosphate excretion is presumably a further contributing factor. The limitation of urinary excretion of ammonium may reflect depressed renal ammoniagenesis, due to the overproduction of keto acids,27 and the failure to produce protons distally to titrate the ammonia. Low-grade loss of bicarbonate is a common finding in distal renal tubular acidosis28 and is partly due to phosphate depletion. Impairment of urinary concentrating ability was a relatively uncommon finding in the alcoholic patients. This defect was due to decreased renal sensitivity to vasopressin, as previously reported29. Since some of these patients consumed large quantities of beer, the prolonged ingestion of large volumes of hypotonic fluids may have caused a decrease in medullary interstitial osmolality, with a consequent reduction in the concentrating ability of the renal medulla.
This study indicates that ethanol is responsible for derangements of monovalent- and divalent-cation metabolism and acid-base homeostasis independently of malnutrition, liver disease, pancreatitis, and other intercurrent illness. There is little doubt that the pathogenesis of hypophosphatemia, hypocalcemia, hypomagnesemia, and hypokalemia is multifactorial. During the acute phase of alcohol withdrawal, the increased plasma insulin concentration, the hyperventilation with respiratory alkalosis, and the elevated plasma epinephrine concentrations probably contributed to hypophosphatemia, hypokalemia, and hypomagnesemia by promoting the movement of ions into cells30,31. However, since the principal defense against the depletion of minerals and electrolytes is the ability to reduce their losses through the kidneys, the tubular dysfunction undoubtedly contributed to the changes in serum ionic composition. In fact, defects in tubular function were more common in the patients who had corresponding serum abnormalities of serum electrolytes.
Approximately one third of the alcoholic patients had metabolic acidosis on admission. Alcoholic ketoacidosis was the most common disturbance, but a wide variety of simple and mixed acid-base disorders were encountered. There is little doubt that the increased generation of keto acids has a major role in the pathogenesis of alcohol-induced metabolic acidosis32,33. However, 55 percent of the patients with metabolic acidosis had impaired renal acidification ability, and 6 of 10 patients in whom metabolic acidosis was not accompanied by respiratory or metabolic alkalosis had abnormally low net excretion of acid despite systemic acidemia. These results suggest that in prolonged, severe ethanol abuse, metabolic acidosis may be induced or aggravated by defects in the renal mechanisms of acid excretion.
Impaired renal acidification and other renal tubular defects are common in patients with chronic liver disease -- e.g., chronic active hepatitis, primary biliary cirrhosis, and alcoholic cirrhosis1,2,3,34. This study, based on a large battery of tests of tubular damage or dysfunction, suggests that the nephrotoxicity of alcohol may be independent of any associated liver disease. The prognostic importance of such renal abnormalities cannot be easily assessed, since the tubular damage is generally mild and transient and does not appear to be associated with progressive renal disease.
Source Information
From the Department of Internal Medicine, University of Udine Medical School, Udine (S.D.M., E.C., A. Bertotti, E.B.); the Department of Internal Medicine, General Hospital, San Vito al Tagliamento (A. Basile); and the Laboratory of Chemistry, Institute of Hygiene, Udine (R.N.) -- all in Italy.
Address reprint requests to Dr. De Marchi at Via Tartagna 39, 33100 Udine, Italy.
References
-hydroxybutyric acid and acetoacetic acid in blood. Biochem J 1962;82:90-96. [Medline]
2-Microglobulin: its significance in the evaluation of renal function. Kidney Int 1987;32:635-641. [Medline]
-hydroxylase. Clin Sci 1984;66:659-663. [Medline]
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
