Nephron 1991;57:137-143
< 1991 S. Karger AG. Basel 0028-2766/91 .'0572-0137S2.75/0
Effect of Acetate on Blood Metabolites and Glucose Tolerance during Haemodialysis in Uraemic Non-Diabetic and Diabetic Subjects Abayomi O. Akanjia, Steven Sacksb “Sheikh Rashid Diabetes Unit, The Radcliffe Infirmary and hThe Renal Unit, The Churchill Hospital, Oxford, UK
Keywords. Acetate • Glucose • Haemodialysis • Ketone bodies • Non-esterified fatty acids • Diabetes mellitus
Introduction During haemodialysis against acetate-containing fluids, a large amount of free acetate enters the circula tion, and may be metabolised [1]. Indeed, acetate oxida tion to carbon dioxide can contribute up to 40% of the total energy expenditure during dialysis [2]. Other studies on acetate metabolism during uraemic haemodialysis reported that (1) acetate may be incorpo rated directly into lipids and thereby contribute to the increased prevalence of atherosclerosis in acetate-dia lysed subjects [3]; (2) acetate follows the anaerobic glyco lytic pathway in a retrograde fashion thereby causing significant elevation of blood lactate [4], and (3) acetate dialysis was associated with reduced levels of glucose and insulin and accelerated ketogenesis [5]. It can there fore be inferred from the foregoing that the oxidation of
acetate increased to meet energy demands while sparing glucose utilisation during glucose-free acetate dialysis [5]. On critical review however, it is obvious that many of the studies did not include appropriate control experi ments and it was difficult to separate any metabolic effect of acetate during dialysis from that due primarily to alkalosis or even prolonged fasting. Such distinction is important in prescribing acetate-or bicarbonate-contain ing fluids for haemodialysis especially in diabetic sub jects who already have gross abnormalities in the inter mediary metabolism of fat and glucose. Also, it is unclear whether some of the morbidity (especially cardiovascular instability and hypotension) associated with acetate hae modialysis could possibly be consequent on the meta bolic end-products of acetate [1]. Furthermore, the effect on glucose tolerance and glycaemic control of frequent exposure to increased blood acetate levels, a source of
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 3:03:47 AM
Abstract. We examined changes in blood concentrations of glucose, acetate and other blood intermediary metabolites as well as the disposal of an intravenous glucose load during successive glucose-free acetate and control bicarbonate haemodialysis in random order, in non-diabetic and diabetic subjects. Plasma acetate levels increased about 10-fold in both the diabetic and non-diabetic subjects during the 1st hour of acetate dialysis. This was accompanied by about 5-fold increase in the levels of the ketone bodies - acetoacetate and 3-hydroxybutyrate. Ketone body levels did not change during bicarbonate dialysis. Additionally, the changes in blood levels of glucose, non-esterified fatty acids, pyruvate, lactate, glycerol and insulin were similar with either bicarbonate or acetate treatment in each subject group. The KCi rate constant of glucose disposal after intravenous glucose (10 g/m 2 body surface area) given after the 1st hour of a 5-hour haemodialysis was similarly unchanged with acetate or bicarbonate dialysis, despite the additional caloric load likely from the high blood levels of acetate obtained during acetate haemodialysis. We conclude that the major metabolic effect of acetate during haemodialysis is an increase in blood ketone body levels. Glucose utilisation, as assessed from the KCl rate constant of glucose disposal or changes in blood lactate and pyruvate levels, was not affected by the high acetate levels attained during dialysis.
A kanji/Sacks
138
Solution A 40 Albumin, g/l Non-esterified fatty acids (as linoleic 1.0 acid), mmol/l 0.5 3-hydroxybutyrate. mmol/l Acetoacetate, mmol/l 0.25 1.0 Lactate, mmol/l 0.2 Pyruvate, mmol/l Glycerol, mmol/l 0.25 Insulin, m U /l 50 10 Glucose, mmol/l
Solution B 40 0.5 0.25 0.125 0.5 0.1 0.1 20 5
extra calories [2], in diabetic subjects is essentially un clear, whether during dialysis or in the interdialysis pe riod. This study therefore attempts to compare the meta bolic changes in stable uraemic patients (diabetic and non-diabetic) on both acetate and bicarbonate dialysis in order to control for the possible effects of changes in anionic metabolism. The effect of the high acetate load on the metabolism of glucose intravenously administered during haemodialysis was also investigated.
Subjects and Methods Subjects
Five male non-diabetic uraemic subjects (aged 50.0 ±13.0 years, mean ± SD, with body mass index 24.5 ±3.7 k g /n r and glycosylated haemoglobin, HBAIc, values 7.9±2.0%), and 5 male diabetic urae mic subjects (aged 53.0±6.6 years with body mass index 23.0± 1.3 kg/m:, HBAIc ll.5±2.7%, duration of diabetes 23.8±9.9 years, all insulin-treated at a daily dose of 0.6±0.3 U/kg body weight) com pleted the studies. They were recruited after voluntary informed consent. All the subjects were in established chronic end-stage renal failure (from chronic glomerulonephritis in the nondiabetic and progressive nephropathy in the diabetic subjects) and had been regularly on haemodialysis for at least 6 months. Their haematocrit was 32.0±2.6% and average predialysis plasma urea and creatinine values were for the non-diabetic subjects 27.6±4.0 mmol/1 and 930±224 nm ol/l, respectively, and for the diabetic subjects 30.2±5.4 mmol/l and 971 ±258 nmol/l, respectively. All were clin ically anephric with creatinine clearance values less than 2 ml/min, and were on thrice weekly haemodialysis at the Renal Unit. Chur chill Hospital, Oxford. None of the subjects was on treatment with steroids or any other drug (except insulin in the diabetics) known to influence glucose metabolism. The studies were approved by the Oxford Area Ethical Committee. The number of patients in the study was relatively small because there was considerable difficulty in recruiting subjects who were in stable glycaemic control, free of intercurrent infections, well enough
to tolerate overnight fasting and who did not suffer hypotensive and hypoglycaemic episodes during glucose-free haemodialysis. In deed, 10 non-diabetic and 8 diabetic subjects were originally recrui ted. but only 5/group could complete the studies. The default rates (due to hypotension or hypoglycaemia) were similar for both acetate and bicarbonate haemodialysis. Procedure
The studies were performed during each patient's routine treat ment programme. The type of dialyser (COBE Centry 2Rx Dialysis Control Unit, Allegro Hollow Fibre. COBE Labs. Lakewood.Colo.. USA), blood flow rate (200 ml/min), dialysate flow' rate (500 m l/ min), duration of each dialysis session (5 h) and frequency of dialysis (thrice per week) were not changed. Vascular access in each patient w'as by forearm arteriovenous fistulae. Blood samples were withdrawn from the arterial end of the fistula, while glucose and heparin infusions were into the venous end. Each subject had isosmolar acetate and bicarbonate dialyses at l-week intervals, in a random fashion. The respective compositions of the dialysates (MaCarthy's Laboratories. Romford, UK) were (mmol/l): acetate dialysate Na 131. K 1.57. Mg 0.5, Ca 1.54, Cl 96,9, glucose 0, acetate 38.0 and bicarbonate dialysate - Na 130, K 2.0, Mg 0.38, Ca 1.5, Cl 105, glucose 0, acetate 2.0, bicarbonate 30.0. We had earlier established that isosmolar acetate and bicarbonate infusates (as in the dialysates above) in non-uraemic diabetic and non-dia betic humans produced identical changes in blood bicarbonate levels and degree of metabolic alkalosis [6], Each patient was anti coagulated during dialysis with 10,000-15,000 units of heparin. The patients arrived in the Renal Unit after a 10- to 12-hour overnight fast (with water drunk as wished) at about 08.30 h on the morning of the study. After a 30-min rest, 2 initial blood samples were taken at 10-min intervals before the dialysis and heparin infusions were commenced. Another specimen was taken at the end of the 1st hour of dialysis and then the study proceeded as for an intravenous glucose tolerance test, after intravenous injection of glucose (10 g/m 2body surface area) w'ith blood sampling at 1,2,5,10, 15,20,30,40.50 and 60 min afterintravenous glucose. The study was usually completed by the end of the 2nd hour of dialysis, after which the patient was fed. Blood pressure and blood glucose (BM-Stix. Boehringer, Mannheim, FRO) measurements were done at 30-min intervals throughout each study period to detect the onset of hypo tension or hypoglycaemia, both conditions usually necessitating discontinuation of the procedure. An initial pilot study was done using two ’dummy plasma’ solutions to assess diffusive losses of plasma constituents into the dialysate. since that was technically impossible to detect with the dialysis set-up for the patients. These dummy plasma solutions A and B. respectively, contained the following (table 1). They were considered to contain concentrations approximate to the levels of the various substrates in diabetic (A) and non-diabetic (B) plasma. It was observed that the fractional losses of these metabolites were identical during acetate and bicarbonate dialysis at the same dialyser settings. The losses were detectable only for glucose in (A) (about 6% at concentrations greater than 7.0 mmol/l) while for the other substrates, albumin and insulin, fractional losses W'ere undetectable. It was therefore assumed during the studies that the arterial plasma values fairly accurately reflected the true values, and that, in the case of glucose, the diffusive loss into the dialysis fluid was not sufficient to significantly affect the qualitative trend of our findings.
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 3:03:47 AM
Tabic 1. Constituents of solutions A and B
M etabolic Effects o f Acetate during Dialysis
139
Table 2. Levels of blood metabolites before and after 60 min dialysis with acetate or bicarbonate in non-diabetic and diabetic subjects (m ean±SD ) Acetate
Bicarbonate
before
after 60 min
before
after 60 min
Nondiabetic (n = 5) Glucose Insulin Lactate Pyruvate Aceto-acetate 3-hydroxybutzyrate NEFA Glycerol Acetate
5.0 ±0.4 12.3 ±3.8 0.64 ±0.38 0.079 ±0.015 0.045 ±0.010 0.034 ±0.019 1.68 ±1.13 0.073 ±0.033 0.18 ± 0 .1 1
4.4 ±0.6 10.2 ±5.0 0.40 ±0.16 0.060 ±0.028 0.068 ± 0 .0 15» 0.153 ±0.100* 2.15 ±0.84* 0.084 ±0.028 1.87 ±0.16*
5.0 ±0.7 11.3 ±2.8 0.55 ±0.08 0.060 ±0.029 0.043 ±0.010 0.048 ±0.051 2.17 ± 1.08 0.081 ±0.012 0.17 ±0.07
4.6 ± 1.0 12.0 ±2.8 0.57 ±0.16 0.055 ±0.028 0.056 ±0.021 0.087 ±0.042 2.29 ±0.71* 0.071 ±0.022 0.29 ±0.19*
Diabetic (n = 5) Glucose Free insulin Lactate Pyruvate Aceto-acetate 3-hydroxy-butyrate NEFA Glycerol Acetate
8.1 ±3.1 14.4 ±4.0 0.92 ±0.46 0.076 ±0.015 0.089 ±0.021 0.153 ±0.130 1.14 ±0.93 0.127 ±0.045 0.28 ±0.06
6.2 ± 1.1 15.4 ±4.9 0.71 ±0.31 0.074 + 0.010 0.132 ±0.028* 0.243 ±0.150* 1.91 ±0.58* 0.097 ±0.052 2.08 ±0.91*
10.0 ±5.3 17.4 ±7.0 0.67 ±0.11 0.089 ±0.033 0.088 ±0.020 0.195 ±0.147 1.21 ±0.88 0.108 ±0.031 0.31 ±0.23
8.0 ±4.3 16.0 ±7.8 0.70 ±0.36 0.079 + 0.031 0.069 ±0.017 0.183 ±0.131 2.32 ±0.44* 0.105 ±0.052 0.38 ±0.26*
Unit of measurement for all the metabolites is mmol/l and for insulin m U /l. *p
< 1991 S. Karger AG. Basel 0028-2766/91 .'0572-0137S2.75/0
Effect of Acetate on Blood Metabolites and Glucose Tolerance during Haemodialysis in Uraemic Non-Diabetic and Diabetic Subjects Abayomi O. Akanjia, Steven Sacksb “Sheikh Rashid Diabetes Unit, The Radcliffe Infirmary and hThe Renal Unit, The Churchill Hospital, Oxford, UK
Keywords. Acetate • Glucose • Haemodialysis • Ketone bodies • Non-esterified fatty acids • Diabetes mellitus
Introduction During haemodialysis against acetate-containing fluids, a large amount of free acetate enters the circula tion, and may be metabolised [1]. Indeed, acetate oxida tion to carbon dioxide can contribute up to 40% of the total energy expenditure during dialysis [2]. Other studies on acetate metabolism during uraemic haemodialysis reported that (1) acetate may be incorpo rated directly into lipids and thereby contribute to the increased prevalence of atherosclerosis in acetate-dia lysed subjects [3]; (2) acetate follows the anaerobic glyco lytic pathway in a retrograde fashion thereby causing significant elevation of blood lactate [4], and (3) acetate dialysis was associated with reduced levels of glucose and insulin and accelerated ketogenesis [5]. It can there fore be inferred from the foregoing that the oxidation of
acetate increased to meet energy demands while sparing glucose utilisation during glucose-free acetate dialysis [5]. On critical review however, it is obvious that many of the studies did not include appropriate control experi ments and it was difficult to separate any metabolic effect of acetate during dialysis from that due primarily to alkalosis or even prolonged fasting. Such distinction is important in prescribing acetate-or bicarbonate-contain ing fluids for haemodialysis especially in diabetic sub jects who already have gross abnormalities in the inter mediary metabolism of fat and glucose. Also, it is unclear whether some of the morbidity (especially cardiovascular instability and hypotension) associated with acetate hae modialysis could possibly be consequent on the meta bolic end-products of acetate [1]. Furthermore, the effect on glucose tolerance and glycaemic control of frequent exposure to increased blood acetate levels, a source of
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 3:03:47 AM
Abstract. We examined changes in blood concentrations of glucose, acetate and other blood intermediary metabolites as well as the disposal of an intravenous glucose load during successive glucose-free acetate and control bicarbonate haemodialysis in random order, in non-diabetic and diabetic subjects. Plasma acetate levels increased about 10-fold in both the diabetic and non-diabetic subjects during the 1st hour of acetate dialysis. This was accompanied by about 5-fold increase in the levels of the ketone bodies - acetoacetate and 3-hydroxybutyrate. Ketone body levels did not change during bicarbonate dialysis. Additionally, the changes in blood levels of glucose, non-esterified fatty acids, pyruvate, lactate, glycerol and insulin were similar with either bicarbonate or acetate treatment in each subject group. The KCi rate constant of glucose disposal after intravenous glucose (10 g/m 2 body surface area) given after the 1st hour of a 5-hour haemodialysis was similarly unchanged with acetate or bicarbonate dialysis, despite the additional caloric load likely from the high blood levels of acetate obtained during acetate haemodialysis. We conclude that the major metabolic effect of acetate during haemodialysis is an increase in blood ketone body levels. Glucose utilisation, as assessed from the KCl rate constant of glucose disposal or changes in blood lactate and pyruvate levels, was not affected by the high acetate levels attained during dialysis.
A kanji/Sacks
138
Solution A 40 Albumin, g/l Non-esterified fatty acids (as linoleic 1.0 acid), mmol/l 0.5 3-hydroxybutyrate. mmol/l Acetoacetate, mmol/l 0.25 1.0 Lactate, mmol/l 0.2 Pyruvate, mmol/l Glycerol, mmol/l 0.25 Insulin, m U /l 50 10 Glucose, mmol/l
Solution B 40 0.5 0.25 0.125 0.5 0.1 0.1 20 5
extra calories [2], in diabetic subjects is essentially un clear, whether during dialysis or in the interdialysis pe riod. This study therefore attempts to compare the meta bolic changes in stable uraemic patients (diabetic and non-diabetic) on both acetate and bicarbonate dialysis in order to control for the possible effects of changes in anionic metabolism. The effect of the high acetate load on the metabolism of glucose intravenously administered during haemodialysis was also investigated.
Subjects and Methods Subjects
Five male non-diabetic uraemic subjects (aged 50.0 ±13.0 years, mean ± SD, with body mass index 24.5 ±3.7 k g /n r and glycosylated haemoglobin, HBAIc, values 7.9±2.0%), and 5 male diabetic urae mic subjects (aged 53.0±6.6 years with body mass index 23.0± 1.3 kg/m:, HBAIc ll.5±2.7%, duration of diabetes 23.8±9.9 years, all insulin-treated at a daily dose of 0.6±0.3 U/kg body weight) com pleted the studies. They were recruited after voluntary informed consent. All the subjects were in established chronic end-stage renal failure (from chronic glomerulonephritis in the nondiabetic and progressive nephropathy in the diabetic subjects) and had been regularly on haemodialysis for at least 6 months. Their haematocrit was 32.0±2.6% and average predialysis plasma urea and creatinine values were for the non-diabetic subjects 27.6±4.0 mmol/1 and 930±224 nm ol/l, respectively, and for the diabetic subjects 30.2±5.4 mmol/l and 971 ±258 nmol/l, respectively. All were clin ically anephric with creatinine clearance values less than 2 ml/min, and were on thrice weekly haemodialysis at the Renal Unit. Chur chill Hospital, Oxford. None of the subjects was on treatment with steroids or any other drug (except insulin in the diabetics) known to influence glucose metabolism. The studies were approved by the Oxford Area Ethical Committee. The number of patients in the study was relatively small because there was considerable difficulty in recruiting subjects who were in stable glycaemic control, free of intercurrent infections, well enough
to tolerate overnight fasting and who did not suffer hypotensive and hypoglycaemic episodes during glucose-free haemodialysis. In deed, 10 non-diabetic and 8 diabetic subjects were originally recrui ted. but only 5/group could complete the studies. The default rates (due to hypotension or hypoglycaemia) were similar for both acetate and bicarbonate haemodialysis. Procedure
The studies were performed during each patient's routine treat ment programme. The type of dialyser (COBE Centry 2Rx Dialysis Control Unit, Allegro Hollow Fibre. COBE Labs. Lakewood.Colo.. USA), blood flow rate (200 ml/min), dialysate flow' rate (500 m l/ min), duration of each dialysis session (5 h) and frequency of dialysis (thrice per week) were not changed. Vascular access in each patient w'as by forearm arteriovenous fistulae. Blood samples were withdrawn from the arterial end of the fistula, while glucose and heparin infusions were into the venous end. Each subject had isosmolar acetate and bicarbonate dialyses at l-week intervals, in a random fashion. The respective compositions of the dialysates (MaCarthy's Laboratories. Romford, UK) were (mmol/l): acetate dialysate Na 131. K 1.57. Mg 0.5, Ca 1.54, Cl 96,9, glucose 0, acetate 38.0 and bicarbonate dialysate - Na 130, K 2.0, Mg 0.38, Ca 1.5, Cl 105, glucose 0, acetate 2.0, bicarbonate 30.0. We had earlier established that isosmolar acetate and bicarbonate infusates (as in the dialysates above) in non-uraemic diabetic and non-dia betic humans produced identical changes in blood bicarbonate levels and degree of metabolic alkalosis [6], Each patient was anti coagulated during dialysis with 10,000-15,000 units of heparin. The patients arrived in the Renal Unit after a 10- to 12-hour overnight fast (with water drunk as wished) at about 08.30 h on the morning of the study. After a 30-min rest, 2 initial blood samples were taken at 10-min intervals before the dialysis and heparin infusions were commenced. Another specimen was taken at the end of the 1st hour of dialysis and then the study proceeded as for an intravenous glucose tolerance test, after intravenous injection of glucose (10 g/m 2body surface area) w'ith blood sampling at 1,2,5,10, 15,20,30,40.50 and 60 min afterintravenous glucose. The study was usually completed by the end of the 2nd hour of dialysis, after which the patient was fed. Blood pressure and blood glucose (BM-Stix. Boehringer, Mannheim, FRO) measurements were done at 30-min intervals throughout each study period to detect the onset of hypo tension or hypoglycaemia, both conditions usually necessitating discontinuation of the procedure. An initial pilot study was done using two ’dummy plasma’ solutions to assess diffusive losses of plasma constituents into the dialysate. since that was technically impossible to detect with the dialysis set-up for the patients. These dummy plasma solutions A and B. respectively, contained the following (table 1). They were considered to contain concentrations approximate to the levels of the various substrates in diabetic (A) and non-diabetic (B) plasma. It was observed that the fractional losses of these metabolites were identical during acetate and bicarbonate dialysis at the same dialyser settings. The losses were detectable only for glucose in (A) (about 6% at concentrations greater than 7.0 mmol/l) while for the other substrates, albumin and insulin, fractional losses W'ere undetectable. It was therefore assumed during the studies that the arterial plasma values fairly accurately reflected the true values, and that, in the case of glucose, the diffusive loss into the dialysis fluid was not sufficient to significantly affect the qualitative trend of our findings.
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 3:03:47 AM
Tabic 1. Constituents of solutions A and B
M etabolic Effects o f Acetate during Dialysis
139
Table 2. Levels of blood metabolites before and after 60 min dialysis with acetate or bicarbonate in non-diabetic and diabetic subjects (m ean±SD ) Acetate
Bicarbonate
before
after 60 min
before
after 60 min
Nondiabetic (n = 5) Glucose Insulin Lactate Pyruvate Aceto-acetate 3-hydroxybutzyrate NEFA Glycerol Acetate
5.0 ±0.4 12.3 ±3.8 0.64 ±0.38 0.079 ±0.015 0.045 ±0.010 0.034 ±0.019 1.68 ±1.13 0.073 ±0.033 0.18 ± 0 .1 1
4.4 ±0.6 10.2 ±5.0 0.40 ±0.16 0.060 ±0.028 0.068 ± 0 .0 15» 0.153 ±0.100* 2.15 ±0.84* 0.084 ±0.028 1.87 ±0.16*
5.0 ±0.7 11.3 ±2.8 0.55 ±0.08 0.060 ±0.029 0.043 ±0.010 0.048 ±0.051 2.17 ± 1.08 0.081 ±0.012 0.17 ±0.07
4.6 ± 1.0 12.0 ±2.8 0.57 ±0.16 0.055 ±0.028 0.056 ±0.021 0.087 ±0.042 2.29 ±0.71* 0.071 ±0.022 0.29 ±0.19*
Diabetic (n = 5) Glucose Free insulin Lactate Pyruvate Aceto-acetate 3-hydroxy-butyrate NEFA Glycerol Acetate
8.1 ±3.1 14.4 ±4.0 0.92 ±0.46 0.076 ±0.015 0.089 ±0.021 0.153 ±0.130 1.14 ±0.93 0.127 ±0.045 0.28 ±0.06
6.2 ± 1.1 15.4 ±4.9 0.71 ±0.31 0.074 + 0.010 0.132 ±0.028* 0.243 ±0.150* 1.91 ±0.58* 0.097 ±0.052 2.08 ±0.91*
10.0 ±5.3 17.4 ±7.0 0.67 ±0.11 0.089 ±0.033 0.088 ±0.020 0.195 ±0.147 1.21 ±0.88 0.108 ±0.031 0.31 ±0.23
8.0 ±4.3 16.0 ±7.8 0.70 ±0.36 0.079 + 0.031 0.069 ±0.017 0.183 ±0.131 2.32 ±0.44* 0.105 ±0.052 0.38 ±0.26*
Unit of measurement for all the metabolites is mmol/l and for insulin m U /l. *p
