- Email: [email protected]
Increased erythrocyte sodium efflux during overfeeding without evidence of mediation by circulating catecholamines or thyroid hormone
Increased Erythrocyte Sodium Efflux During Overfeeding Without Evidence of Mediation by Circulating Catecholamines or Thyroid Hormone Bjijrn Fagerberg,
Hans He&z,
Olof Jonsson,
G&an
Lindstedt,
Jacob Nauckr,
Ulla Nilsson, Thomas
Hedner,
and Ove Andersson
Ten slightly obese middle-aged men were instructed to increase their energy intake 25% during a period of 1 week, which was preceded by a control period of seven days. Body weight increased by 0.67 kg (SD 0.601 indicating good compliance with the regimen. Transmembrane sodium fluxes were determined with the use of =Na. The pre-diet erythrocyte sodium content was 9.7 mmol/L (SD 0.6) decreasing to 6.9 mmol/L (SD 1 .l I (P < 0.05) during overfeeding. The Na-efflux rate constant increased from 0.40 h-’ to 0.54 h-’ (P c 0.05). Urinary excretion of catecholamines and concentrations of catecholamines and insulin in plasma and of thyroxine, triiodothyronine, and reverse T3 in serum did not change. Thus, overfeeding seems to enhance the total Na efflux in erythrocytes from slightly obese men. There were no measurable changes in thyroid hormone or catecholamine levels leaving the regulatory mechanisms unexplained.
D
URING experimental overfeeding in normal men there is an adaptive increase in energy expenditure called dietary thermogenesis, explaining why the observed weight gain often is less than would be predicted.’ The mechanism of this enhanced metabolic rate is still unknown, although it has been shown that overnutrition is associated with an increase both in serum 3, 5, 3’-triiodothyronine concentration’,’ and sympathetic nervous system activity3,4 as well as in plasma insulin concentrati0n.j The Na+ K’ pump, a constituent of all cells, appears to contribute significantly to cellular thermogenesis and energy balance6-* although this is a matter of dispute.’ The fact that thyroid hormones,6 catecholamines,” and insulin” have been shown to have specific effects on the Na+ K’ pump provides a possible explanation: Overfeeding might cause an elevation of the Na’ K’-pump activity and cellular thermogenesis, mediated by an increase in one or several of the regulatory systems mentioned above. The composition of the diet may be of importance because refined carbohydrates but not fat seem to increase indices of sympathetic nervous activity.‘2Y’3It has also been shown that increased energy intake in the form of sucrose supplement leads to an increase in Na+ K+-ATPase-mediated ion transport in liver and muscle in control mice but not in ob/ob mice.14 From the Department of Internal Medicine I. the Department of Nephrology, the Department of Surgery I, the Department of Urology. the Department of Clinical Nutrition, the Department of Clinical Pharmacology, and the Department of Clinical Chemistry, Sahlgrenska Hospital, University of Giiteborg, Giiteborg. Sweden. This study was supported by the Giiteborg Medical Society, the Swedish Medical Research Council (no 58410406-1, B 84-19x0.575-05A). the National Society againsi Heart and Lung Disease, and The Swedish Society of Medical Sciences. Address reprint requests to Dr Bjijrn Fagerberg, Medical Department I. Sahlgrenska sjukhuset. S-413 45 Gijteborg, Sweden. 0 I984 by Grune & Stratton, Inc.
994
We have now studied the effect of overfeeding of slightly obese men on intracellular electrolytes and sodium fluxes in erythrocytes and if there are any concomitant changes in plasma insulin or in thyroid and sympathoadrenal activity. MATERIAL
AND
METHODS
Twelve middle-aged (42 to 58 years), slightly obese but otherwise normal men were recruited through advertisements in a local newspaper. The amount overweight was 15% to 35% in relation to a (Table I). The subjects were studScandinavian ideal population,” ied on an outpatient basis, thus, occupationally fully active. After an initial examination and informed consent the subjects entered the study, which consisted of three periods: basal, control, and diet periods each of seven days. The volunteers were told not to change their habits concerning alcohol consumption or smoking. The dietary instructions were given by a dietitian. In the basal period the subjects were instructed not to change their food intake, which they recorded for seven days. During the control period the diet was formalized on the basis of the earlier diet records aiming at an unchanged body mass. In the diet period the volunteers increased their energy intake 25% in comparison with the control period. Six volunteers were randomized to overfeeding with sugar, the other six to overnutrition with butter. Diet records were obtained for all seven days in this final period. In each of the control and diet periods the subjects delivered two 24-hour urine specimens for determination of sodium, noradrenaline, and adrenaline. At the end of each period, the following protocol was applied with the exception of sustained handgrip and erythrocyte ion studies not being performed in the basal period: the subject arrived at 8 AM at the clinic after an overnight fast. The body mass was measured on a balance scale (Stathmos, Sweden) with the volunteer dressed in trousers. A catheter was inserted in the left antecubital vein and the subject then rested for ten minutes in a sitting position. Next, the device for performing sustained handgrip was demonstrated. It consisted of a balloon and a manometer (Vigorimeter, Gebr. Martin. Tuttlingen, Germany). The volunteer was asked to squeeze the balloon as hard as he could. He was then instructed to squeeze the balloon to keep the needle at 30% of his previous maximum value during three minutes. At the end of this isometric exercise, blood was drawn for determination of plasma noradenaline. The subject was then placed in a supine position, resting for 30 minutes during which blood pressure and heart rate were measured every five minutes in a conventional manner with a cuff (16 x 36 cm) connected to a Metabolism,
Vol 33, No 1 1 (November), 1984
SODIUM EFFLUX DURING OVERFEEDING
995
Table 1. Subject Data BZWXll
87.8
Body mass (kg) t&sodium
(mmol/L/24
+ 8.8
h)
Control
87.6
Overfeeding
+ 6.6
88.3
& 6.5t
199.2
? 70.7
190.5
t 47.4
41.1
k 14.7
35.0
& 16.7
t&adrenaline (nmol/L/24
h)
tU-noradrenaline
lnmol/L/24
321 + 105
301 + 148
96.6 k 8.9
1.77 * 0.59 2.37 + 1.12 97.1 -+ 10.7
1.66 + 0.41 2.25 + 0.89 94.9 k 7.8
1.82 i 0.2 0.22 + 0.07 18.1 t 17.5
1.80 + 0.24 0.21 2 0.1 23.0 + 24.3
1.75 + 0.19 0.21 i 0.04 10.7 zk 4.8
h)
P-noradrenaline (nmol/L) Resting
1.77 f 0.6
l-land-grip Thyroxine (nmol/L) Triidothyronine (nmol/L) Reverse T3 (nmol/L) Plasma insulin (mu/L)
120, and 240 minutes the samples were put on a column for separation of the erythrocytes according to Henningsen.*’ In this procedure an aliquot of 0.5 mL was used and the erythrocytes were separated from the solution on a Sephadex G 50 (fine) column at 22 “C (length 20 cm. diameter 1.6 cm, flow I mL/min), using a buffer (I vol of 0.15 mol/L TRIS with IO vol of 0.153 mol/L NaCl pH 7.4) as effluent. One fraction of the erythrocytes was taken and centrifuged for two minutes at 2000 g at 4 “C and the erythrocyte volume fraction was measured. After separation, 200 rL of the sample was collected in duplicate for calculation of 2ZNa activity in a gamma counter and the activity in 200 PL incubation medium was counted simultaneously. It has been shown that the ‘*Na uptake can be described by one monoexponential function (Correlation coefficient = 0.9814 t 0.0024, n = IO**). This monoexponential uptake can be written as: Y =
E + y (I ~ emk’)
Blood pressure (mmtig) Systolic Diastolic Heart rate (beats/min) Body mass, 24-hour
138.3 + 21.7 135.1 t 17.8 138.6? 16.3 90.6 t 87.6 88.5 f 8.2 90.8 + 10.2 63.7 + 10.3 65.3 + 10.1 63.4 + 7.1 urinary excretion of sodium, adrenaline, and
noradrenaline; plasma noradrenaline (resting and during sustained handgrip for three minutes); serum levels of triiodothyronine, reverse T3, thyroxine: plasma insulin; blood pressure; and heart rate in ten men at the end of the basal, control, and diet periods. Values given are mean + SD. tP < 0.0 1 compared with control values.
mercury manometer. Blood samples for assay of intraerythrocyte sodium and potassium, transmembrane fluxes of sodium, insulin. thyroid hormones, and resting noradrenaline were then drawn. Serum thyroxine (T4), 3, 5, 3’-triiodothyronine (T3) and 3, 3’, 5’-triiodothyronine (reverse T3) were analyzed by radioimmunoassay (RIA).i6 Plasma noradrenaline was determined by high-pressure liquid chromatography as earlier described.” Plasma insulin was determined by Phadebas Insulin Test (Pharmacia AB, Uppsala, Sweden). In order to preserve the urinary catecholamines, the subjects cohected their specimens in cans containing 16 mL 5 mol/L HCI. Urinary concentrations of adrenaline and noradrenaline were assayed according to a method originally described by von Euler and Lishajko.‘” Urinary sodium was measured with flame photometry. Erythrocyte sodium content was determined according to the method previously described.” Venous blood was collected in heparinized tubes and the blood samples were immediately placed in an ice-water bath. After centrifugation (4 “C, 3000 g, five minutes), I vol of packed cells was washed twice with 3 vol of MgCI (96.7 mmol/L) buffered to pH 7.2 by isotonic TRIS buffer. An aliquot of 50 rL of washed erythrocyte suspension was added to a tube containing 200 FL concentrated nitric acid and the rest of the washed erythrocyte suspension was used for determination of erythrocyte volume fraction. Sodium content in the samples was analyzed in triplicate using a flame photometer. Correction was made for trapped sodium in the washing solution. Determination of so&urn influx and et&x was done as described by Herlitz et al.“’ Blood samples were collected in heparin tubes and were centrifuged at 3000 g for five minutes. After removal of plasma and huffy coat, 2 mL of packed erythrocytes was washed once with 2 mL of the incubation solution (Dulbecco’s phosphate buffered saline also containing 8.9 mmol/L of D-glucose and 50 g/L of human albumin) and then again centrifuged. One mL of blood was incubated in 2 mL incubation solution during shaking in a water bath at 37 “C. Radioactive sodium (‘INa) was added with an initial specific activity of approximately 0.19 MBq x mall’. After incubation for 15, 60,
where m denotes influx, k: the rate constant for cellular efflux, and E. the rapid initial uptake of 22Na, which in erythrocytes is close to zero. Estimates form, k, and E that make this formula best, describe the observed uptake values for ‘“Na as calculated according to the least-square method.** Sodium influx is expressed in mmol x 1-I x h-’ and the rate is constant for sodium efflux in h- ‘. Exchangeable IeNa was calculated from the computed uptake curve as the erythrocyte sodium content at infinite time. This method for assessment of sodium influx and efflux yields similar values as separate determinations of sodium uptake and washout in red blood cells.24~*5 The study was approved by the Ethical Committee of the University of Giiteborg. The paired t test and linear correlation were used for statistical analysis. RESULTS
Two volunteers were excluded due to noncompliance. The remaining subjects had an age of 51.2 f 5.6 (mean * SD) years, and a body mass of 87.8 kg (SD 6.8). The subjects (n = 5) who were prescribed sucrose overfeeding showed a pre-diet energy intake of 2020 kcal/d (SD 784) (8.5 MJ) per day rising to 2500 kcal (SD 758) (10.5 MJ) according to dietary records. In the other group (n = 5) the basal energy intake was 1650 kcal/d (SD 380) (6.9 MJ) and 2220 kcal/d (SD 580) (9.2 MJ) during the diet period with fat supplementation. In the presentation of the results the subjects are pooled from the following reasons: first, the two groups seemed to be overfed to a similar degree as shown by the diet records and the almost identical increase in body mass. Neither were there any significant differences in sodium excretion. Second, there were no indications of differing responses in catecholamines, thyroid hormones, or plasma insulin to overfeeding with either fat or sugar. Third, in both groups there were similar changes in intraerythrocyte sodium content and ion fluxes (Table 2). During seven days of overfeeding, body mass increased by 0.67 kg (SD 0.6) (Tables 1, 2). The 24-hour urinary sodium excretion did not change significantly between the study periods (Table I). There were no significant changes in urinary output or in
996
FAGERBERG ET AL
Table 2.
Subject (Diet) 1
Individual Chanaes in Bodv Mass, lntraervthrocvte Na
bnmol/L)
Sodium Content
and ion Fluxes
Na-influx fmmol x-’ 1 x h-‘1
Na-effluX rate constant fh-‘)
K ~mmol/Ll
Period
aodv Mass
Total
Control
95.2
10.1
5.8
2.5
0.45
Diet
95.6
9.1
4.4
3.4
0.76
98.6
Control
88.0
11.7
6.1
3.8
0.62
86.4
Diet
88.5
10.2
7.6
2.8
0.37
82.5
Control
81.5
9.2
6.3
3.1
0.50
91.0
Diet
82.5
9.0
6.0
3.4
0.57
96.4
4
Control
80.2
8.1
7.1
2.3
0.35
94.2
(F)
Diet
80.2
8.8
5.0
2.1
0.45
94.9
Control
90.0
9.9
7.1
1.9
0.28
91.8
Diet
91.0
9.5
6.3
2.1
0.35
87.4
Control
81.8
8.9
7.8
1.4
0.19
97.1
Diet
83.4
7.7
4.2
2.7
0.68
93.9
Control
100.3
9.6
5.5
2.5
0.48
90.4
Diet
(CH) 2 (F) 3 (CH)
5 (CH) 6 (F) 7 (CH) 8 (F) 9 (CH)
Exchangeable
100.0
9.9
5.0
3.2
0.65
so.2
Control
86.2
9.9
7.3
2.0
0.29
98.6
Diet
86.8
7.5
5.3
2.7
0.52
92.3
Control
91.0
ND
ND
ND
ND
ND
Diet
92.5
(7.3)
(6.7)
(2.7)
(0.88)
(94.2)
(0.211
(86.6)
10
Control
81.9
(9.1)
(4.9)
(2.81
(F)
Diet
82.3
ND
ND
ND
Mean SD
91.0
ND
ND
Control
87.6
+ 6.6
9.7 ? 0.8
6.6 + 0.8
2.4 k 0.7
0.40
i 0.14
92.1
+ 3.
Diet
88.3
f 6.5t
8.9 k 1.1’
5.5 + 1.1.
2.8
0.54
k 0.15’
92.1
* 4.
k 0.5
Body mass, intraerythrocyte sodium content (total and exchangeable), potassium content, sodium influx, and sodium efflux rate constant before and after overfeeding with 25 energy % fat IF) or carbohydrates (CH). ND - not done. l< 0.05 compared with control values. tc0.01
compared with control values.
plasma concentrations of catecholamines or in serum levels of thyroxine, T3, or reverse T3; nor were there any changes in blood pressure or heart rate (Table 1). Plasma noradrenaline after sustained handgrip and insulin levels showed no significant changes when the diet period was compared to baseline (Table 1). The sodium content (total and exchangeable) of the red blood cells, however, decreased significantly (P < 0.05, respectively) with high energy intake (Table 2). Intraerythrocyte potassium concentration did not change, and no difference in the erythrocyte water content was noted. The sodium efflux rate constant increased significantly after overfeeding compared to the control period (P < 0.05) whereas no significant change in sodium influx was noted. No significant correlation was found between weight gain and increase in efflux rate constant during overfeeding (r = 0.57, P < 0.1). DISCUSSION
In outpatient studies comprising changes in diet it is often difficult for the subjects to adhere strictly to the protocol. Furthermore, it is complicated to evaluate the adherence to the regimen and a general experience is that dietary records often underestimate the energy
intake. This is supported by the results of the present study showing an unexpected low energy intake during basal conditions in both groups. In addition, it has been shown that body mass is a poor indicator of overfeeding with a high interindividual variability probably as a reflexion of dietary thermogenesis.’ In our study, there was no significant change in any parameters during the control period but a significant increase in energy intake according to the dietary records during the diet period. Body mass increased in most patients. In conclusion, we believe that the volunteers were overfed according to the study plan. The present study demonstrates that a moderate increase in energy intake during 1 week is associated with an elevation of the erythrocyte efflux rate constant and a reduction of the red blood cell sodium content. Because we have only measured the total efflux, we are not able to determine whether the increase in efflux rate constant could be ascribed to changes in Na+ KC-pump activity or Na+ K’ cotransport. It has, however, been shown in experiments with different transport inhibitors that the ouabainsensitive Na+ K+ pump covers about 75% of the total sodium efflux while Na’ K+ cotransport only contributes to about 12%~~~It therefore seems reasonable to assume that the present increase in sodium efflux is
SODIUM
997
EFFLUX DURING OVERFEEDING
related to an elevation of the Na+ K+-pump activity but a concomitant influence of the cotransport can of course not be excluded. There are earlier studies on the effects of different caloric intake levels on the Na+ K’ pump. It has been found that five days of sucrose overfeeding causes an increase in Na” K’-ATP-ase mediated ion transport in liver slices and intact soleus muscles in normal mice.14 Using an oubain-inhibitable ATP-ase assay on human erythrocyte vesicles Mir et al found an increase in sodium-potassium ATP-ase activity 60 minutes after a meal with a high interindividual variability.26 DeLuise and co-workers on the other hand, found no effects on the number or activity of Na’ K’-ATP-ase units in erythrocytes measured by (3 H) oubain binding and ?ubidium uptake after acute feeding on a hypocaloric diet, fasting, and refeeding in humans.*’ The discrepancy between these studies illustrates not only the problem of differing methods but also the question of whether the erythrocyte is a representative model for transmembrane ion fluxes in all tissues. When comparing studies in the Na’ K’ ATP-ase activity in liver tissue on one hand and in red blood cells on the other, conflicting results have been reported concerning human obesity28.29as well as hyperthyroidism.30,3’ ,4gainst the background that the sodium-potassium pump activity represents most of the total sodium efflux and that the pump in different tissues (muscle, liver) has been shown to be influenced by insulin,” thyroid hormone,6 and catecholamines” we found it interesting to study the dietary effects on these hor-
mones especially because overfeeding is associated with an increase in circulating insulin’ and T32 with decrease in reverse T3.* The present study, however, showed no such changes concerning plasma insulin, T3, T4, or reverse T3. Young and Landsberg have reported that noradrenaline turnover rises in response to overfeeding in rats.“’ Recent short-term studies on humans have shown that overfeeding increased noradrenaline turnover,4 peak standing plasma noradrenaline, and resting metabolic rate.3 We, however, found no evidence of an increase in sympathetic nervous system activity because there were no significant changes in either urinary noradrenaline output or plasma noradrenaline during supine rest or isometric exercise. Nor were there any changes in heart rate, systolic or diastolic blood pressure, as were found in the studies mentioned above.3,4 The urinary adrenaline excretion was not affected by increasing the energy intake. The fact that we could not detect any differences in insulin and thyroid hormone levels or in indices of sympathetic nervous system activity does not exclude the possibility that there may be such a differing response to overfeeding with sugar or fat because our study groups were small. In conclusion, we have found that moderate overfeeding induces an increased sodium efflux and a decreased sodium content in erythrocytes from slightly obese men. Because no estimation of expenditure was obtained our findings can not be related to thermogenesis.
REFERENCES I Danforth Jr E, Burger AG, Goldman RF, et al: Thermogenesis during weight gain, in Bray G (ed): Recent Advances in Obesity Research: II. Proceedings of the 2:nd International Congress on Obesity. London, Newman Publishing Ltd, 1978, pp 229-236 2 Danforth Jr E, Horton ES, O’Connell M, et al: Dietary induced alterations in thyroid metabolism during overnutrition. J Clin Invest 64:1336-1347, 1979 3. Welle S, Campbell RG: Stimulation of thermogenesis by carbohydrate overfeeding. Evidence against sympathetic nervous system mediation. J Clin Invest 7 I:9 16-925, 1983 4. O’Dea K, Esler M, Leonard P, et al: Noradrenaline turnover during under- and over-eating in normal weight subjects. Metabolism 3 I :896-899, I982 5. Sims EAH, Danforth Jr E, Horton ES, et al: Endocrine and metabolic effects of experimental obesity in man. Ret Prog Horm Res 291457496, 1973 6. Smith TJ, Edelman IS: The role of sodium transport in thyroid thermogenesis. Fed Proc 38:215&2153, 1979 7. Ismail-Beigi F, Edelman IS: The mechanism of the calorigenic action of thyroid hormone. Stimulation of Na’, K+-activated adenosine-triphosphatase activity. J Gen Physiol 57:710-722, 1971 8. Whittam R: Active cation transport as a pace-maker of respiration. Nature 191:603-604, 1961 9. Himms-Hagen J: Cellular thermogenesis. Ann Rev Physiol 38:315-351, 1976
10. Clausen T, Flatman J: The effect of catecholamines transport and membrane potential in rat soleus muscle. 270:383-414, 1977
on Na-K J Physiol
11, Clausen T. Kohn PG: The effect of insulin on the transport of sodium and potassium in rat soleus muscle. J Physiol 265:19--42. 1977 12. Welle S, Lilavivat U, Campbell RG: Thermic effect of feeding in man: Increased plasma norepinephrine levels following glucose but not protein or fat consumption. Metabolism 30:953-968, 1981 13. Landsberg L, Young JB: Diet and the sympathetic nervous system: relationship to hypertension. Int J Obesity 5 (suppl) 1:7991. 1981 14. Flier JS, Usher P, DeLuise M: Effect of sucrose overfeeding on Na, K-ATP-ase-mediated 86Rb uptake in normal and ob/ob mice. Diabetes 30:975-978, 1981 15. Lindberg W, Natwig H, Rygh A, et al: Hoyde-og vektundersogelser hos voksne menn og kvinner. Forslag till nye norske hoydevektnormer. Tidskr Norske Laegefor 76:361--368, 1956 16. Blomqvist N, Lindstedt G, Lundberg P-A, et al: Noinhibition by Li+ of thyroxine monodeiodination to 3,5,3’-triiodothyronine and 3,3’.5’-triiodothyronine. Clin Chim Acta 79:4577464. 1977 17. Hjemdahl P, Daleskog M, Kahan T: Determination of plasma catecholamines by high performance liquid chromatography with
998
electrochemical detection: Comparison with a radioenzymatic method. LifeSci 25131-138, 1979 18. v Euler US, Lishajko F: The estimation of catecholamines in urine. Acta Physiol Stand 45:122-131, 1959 19. And& L, Piros S, Hansson L, et al: Different haemodynamic reaction patterns during noise exposure in normotensive subjects with and without heredity for essential hypertension. Clin Sci 63:371-374, 1982 20. Herlitz H, Jonsson 0, Elmfeldt D, et al: Intraerythrocyte sodium content and transmembrane fluxes in patients treated with diuretics or beta-blockers. Acta Pharmacol Toxic01 54 (Suppl) 1:3741,1984 21. Henningsen NC, Nelson D: Net influx and efflux of 22Na in erythrocytes from normotensive offspring of patients with essential hypertension. Acta Med Stand 210:85-91, 1981 22. Jonsson 0, Herlitz H, Gudmundsson 0, et al: Fluxes of sodium and potassium in human erythrocytes: Determination based on computer analyses of isotope uptake curves. Acta Physiol Stand 121:93-95,1984 23. Naucler J, Jonsson 0, Lundstam S, et al: Transmembrane fluxes of potassium in dog kidney slices. A quantitation of the effect of warm ischemia. Eur Surg Res 9:384-396, 1977 24. Mir MA, Bobinski H: Altered membrane sodium transport
FAGERBERG ET AL
and the presence of a plasma oubain-like inhibitory factor in acute myeloid leucemia. Clin Sci Mol Med 48:213-218, 1975 25. Walter U, Distler A: Abnormal sodium efflux in erythrocytes of patients with essential hypertension. Hypertension 4:205-210, 1982 26. Mir MA, Charalombous BM, Morgan K: Erythrocyte sodiurn-potassium-ATP-ase and obesity. N Engl J Med 306:809-810, 1982 27. DeLuise M, Izumo H, Grace EE, et al: Effect of diet upon the erythrocyte Na, K pump. J Endocrinol Metab 56:739-743, 1983 28. Bray GA, Kral JG, Bjorntorp P: Hepatic sodium-potassium dependent ATP-ase in obesity. N Engl J Med 304:1580-l 582, 198 1 29. De Luise M, Blackburn GL, Flier JS: Reduced activity of the red cell sodium-potassium pump in human obesity. N Engl J Med 303:1017-22, 1980 30. Gambert SR, Ingbar SH, Hagen TC: Interaction of age and thyroid hormone status on Nat K+ ATP-ase in rat renal cortex and liver. Endocrinology 108:27-30, 1981 31. Cole CH, Waddel RW: Alteration in intracellular sodium concentration and oubain-sensitive ATP-ase in erythrocytes from hyperthyroid patients. J Clin Endocrinol Metab 42:1056-1063, 1976
Hans He&z,
Olof Jonsson,
G&an
Lindstedt,
Jacob Nauckr,
Ulla Nilsson, Thomas
Hedner,
and Ove Andersson
Ten slightly obese middle-aged men were instructed to increase their energy intake 25% during a period of 1 week, which was preceded by a control period of seven days. Body weight increased by 0.67 kg (SD 0.601 indicating good compliance with the regimen. Transmembrane sodium fluxes were determined with the use of =Na. The pre-diet erythrocyte sodium content was 9.7 mmol/L (SD 0.6) decreasing to 6.9 mmol/L (SD 1 .l I (P < 0.05) during overfeeding. The Na-efflux rate constant increased from 0.40 h-’ to 0.54 h-’ (P c 0.05). Urinary excretion of catecholamines and concentrations of catecholamines and insulin in plasma and of thyroxine, triiodothyronine, and reverse T3 in serum did not change. Thus, overfeeding seems to enhance the total Na efflux in erythrocytes from slightly obese men. There were no measurable changes in thyroid hormone or catecholamine levels leaving the regulatory mechanisms unexplained.
D
URING experimental overfeeding in normal men there is an adaptive increase in energy expenditure called dietary thermogenesis, explaining why the observed weight gain often is less than would be predicted.’ The mechanism of this enhanced metabolic rate is still unknown, although it has been shown that overnutrition is associated with an increase both in serum 3, 5, 3’-triiodothyronine concentration’,’ and sympathetic nervous system activity3,4 as well as in plasma insulin concentrati0n.j The Na+ K’ pump, a constituent of all cells, appears to contribute significantly to cellular thermogenesis and energy balance6-* although this is a matter of dispute.’ The fact that thyroid hormones,6 catecholamines,” and insulin” have been shown to have specific effects on the Na+ K’ pump provides a possible explanation: Overfeeding might cause an elevation of the Na’ K’-pump activity and cellular thermogenesis, mediated by an increase in one or several of the regulatory systems mentioned above. The composition of the diet may be of importance because refined carbohydrates but not fat seem to increase indices of sympathetic nervous activity.‘2Y’3It has also been shown that increased energy intake in the form of sucrose supplement leads to an increase in Na+ K+-ATPase-mediated ion transport in liver and muscle in control mice but not in ob/ob mice.14 From the Department of Internal Medicine I. the Department of Nephrology, the Department of Surgery I, the Department of Urology. the Department of Clinical Nutrition, the Department of Clinical Pharmacology, and the Department of Clinical Chemistry, Sahlgrenska Hospital, University of Giiteborg, Giiteborg. Sweden. This study was supported by the Giiteborg Medical Society, the Swedish Medical Research Council (no 58410406-1, B 84-19x0.575-05A). the National Society againsi Heart and Lung Disease, and The Swedish Society of Medical Sciences. Address reprint requests to Dr Bjijrn Fagerberg, Medical Department I. Sahlgrenska sjukhuset. S-413 45 Gijteborg, Sweden. 0 I984 by Grune & Stratton, Inc.
994
We have now studied the effect of overfeeding of slightly obese men on intracellular electrolytes and sodium fluxes in erythrocytes and if there are any concomitant changes in plasma insulin or in thyroid and sympathoadrenal activity. MATERIAL
AND
METHODS
Twelve middle-aged (42 to 58 years), slightly obese but otherwise normal men were recruited through advertisements in a local newspaper. The amount overweight was 15% to 35% in relation to a (Table I). The subjects were studScandinavian ideal population,” ied on an outpatient basis, thus, occupationally fully active. After an initial examination and informed consent the subjects entered the study, which consisted of three periods: basal, control, and diet periods each of seven days. The volunteers were told not to change their habits concerning alcohol consumption or smoking. The dietary instructions were given by a dietitian. In the basal period the subjects were instructed not to change their food intake, which they recorded for seven days. During the control period the diet was formalized on the basis of the earlier diet records aiming at an unchanged body mass. In the diet period the volunteers increased their energy intake 25% in comparison with the control period. Six volunteers were randomized to overfeeding with sugar, the other six to overnutrition with butter. Diet records were obtained for all seven days in this final period. In each of the control and diet periods the subjects delivered two 24-hour urine specimens for determination of sodium, noradrenaline, and adrenaline. At the end of each period, the following protocol was applied with the exception of sustained handgrip and erythrocyte ion studies not being performed in the basal period: the subject arrived at 8 AM at the clinic after an overnight fast. The body mass was measured on a balance scale (Stathmos, Sweden) with the volunteer dressed in trousers. A catheter was inserted in the left antecubital vein and the subject then rested for ten minutes in a sitting position. Next, the device for performing sustained handgrip was demonstrated. It consisted of a balloon and a manometer (Vigorimeter, Gebr. Martin. Tuttlingen, Germany). The volunteer was asked to squeeze the balloon as hard as he could. He was then instructed to squeeze the balloon to keep the needle at 30% of his previous maximum value during three minutes. At the end of this isometric exercise, blood was drawn for determination of plasma noradenaline. The subject was then placed in a supine position, resting for 30 minutes during which blood pressure and heart rate were measured every five minutes in a conventional manner with a cuff (16 x 36 cm) connected to a Metabolism,
Vol 33, No 1 1 (November), 1984
SODIUM EFFLUX DURING OVERFEEDING
995
Table 1. Subject Data BZWXll
87.8
Body mass (kg) t&sodium
(mmol/L/24
+ 8.8
h)
Control
87.6
Overfeeding
+ 6.6
88.3
& 6.5t
199.2
? 70.7
190.5
t 47.4
41.1
k 14.7
35.0
& 16.7
t&adrenaline (nmol/L/24
h)
tU-noradrenaline
lnmol/L/24
321 + 105
301 + 148
96.6 k 8.9
1.77 * 0.59 2.37 + 1.12 97.1 -+ 10.7
1.66 + 0.41 2.25 + 0.89 94.9 k 7.8
1.82 i 0.2 0.22 + 0.07 18.1 t 17.5
1.80 + 0.24 0.21 2 0.1 23.0 + 24.3
1.75 + 0.19 0.21 i 0.04 10.7 zk 4.8
h)
P-noradrenaline (nmol/L) Resting
1.77 f 0.6
l-land-grip Thyroxine (nmol/L) Triidothyronine (nmol/L) Reverse T3 (nmol/L) Plasma insulin (mu/L)
120, and 240 minutes the samples were put on a column for separation of the erythrocytes according to Henningsen.*’ In this procedure an aliquot of 0.5 mL was used and the erythrocytes were separated from the solution on a Sephadex G 50 (fine) column at 22 “C (length 20 cm. diameter 1.6 cm, flow I mL/min), using a buffer (I vol of 0.15 mol/L TRIS with IO vol of 0.153 mol/L NaCl pH 7.4) as effluent. One fraction of the erythrocytes was taken and centrifuged for two minutes at 2000 g at 4 “C and the erythrocyte volume fraction was measured. After separation, 200 rL of the sample was collected in duplicate for calculation of 2ZNa activity in a gamma counter and the activity in 200 PL incubation medium was counted simultaneously. It has been shown that the ‘*Na uptake can be described by one monoexponential function (Correlation coefficient = 0.9814 t 0.0024, n = IO**). This monoexponential uptake can be written as: Y =
E + y (I ~ emk’)
Blood pressure (mmtig) Systolic Diastolic Heart rate (beats/min) Body mass, 24-hour
138.3 + 21.7 135.1 t 17.8 138.6? 16.3 90.6 t 87.6 88.5 f 8.2 90.8 + 10.2 63.7 + 10.3 65.3 + 10.1 63.4 + 7.1 urinary excretion of sodium, adrenaline, and
noradrenaline; plasma noradrenaline (resting and during sustained handgrip for three minutes); serum levels of triiodothyronine, reverse T3, thyroxine: plasma insulin; blood pressure; and heart rate in ten men at the end of the basal, control, and diet periods. Values given are mean + SD. tP < 0.0 1 compared with control values.
mercury manometer. Blood samples for assay of intraerythrocyte sodium and potassium, transmembrane fluxes of sodium, insulin. thyroid hormones, and resting noradrenaline were then drawn. Serum thyroxine (T4), 3, 5, 3’-triiodothyronine (T3) and 3, 3’, 5’-triiodothyronine (reverse T3) were analyzed by radioimmunoassay (RIA).i6 Plasma noradrenaline was determined by high-pressure liquid chromatography as earlier described.” Plasma insulin was determined by Phadebas Insulin Test (Pharmacia AB, Uppsala, Sweden). In order to preserve the urinary catecholamines, the subjects cohected their specimens in cans containing 16 mL 5 mol/L HCI. Urinary concentrations of adrenaline and noradrenaline were assayed according to a method originally described by von Euler and Lishajko.‘” Urinary sodium was measured with flame photometry. Erythrocyte sodium content was determined according to the method previously described.” Venous blood was collected in heparinized tubes and the blood samples were immediately placed in an ice-water bath. After centrifugation (4 “C, 3000 g, five minutes), I vol of packed cells was washed twice with 3 vol of MgCI (96.7 mmol/L) buffered to pH 7.2 by isotonic TRIS buffer. An aliquot of 50 rL of washed erythrocyte suspension was added to a tube containing 200 FL concentrated nitric acid and the rest of the washed erythrocyte suspension was used for determination of erythrocyte volume fraction. Sodium content in the samples was analyzed in triplicate using a flame photometer. Correction was made for trapped sodium in the washing solution. Determination of so&urn influx and et&x was done as described by Herlitz et al.“’ Blood samples were collected in heparin tubes and were centrifuged at 3000 g for five minutes. After removal of plasma and huffy coat, 2 mL of packed erythrocytes was washed once with 2 mL of the incubation solution (Dulbecco’s phosphate buffered saline also containing 8.9 mmol/L of D-glucose and 50 g/L of human albumin) and then again centrifuged. One mL of blood was incubated in 2 mL incubation solution during shaking in a water bath at 37 “C. Radioactive sodium (‘INa) was added with an initial specific activity of approximately 0.19 MBq x mall’. After incubation for 15, 60,
where m denotes influx, k: the rate constant for cellular efflux, and E. the rapid initial uptake of 22Na, which in erythrocytes is close to zero. Estimates form, k, and E that make this formula best, describe the observed uptake values for ‘“Na as calculated according to the least-square method.** Sodium influx is expressed in mmol x 1-I x h-’ and the rate is constant for sodium efflux in h- ‘. Exchangeable IeNa was calculated from the computed uptake curve as the erythrocyte sodium content at infinite time. This method for assessment of sodium influx and efflux yields similar values as separate determinations of sodium uptake and washout in red blood cells.24~*5 The study was approved by the Ethical Committee of the University of Giiteborg. The paired t test and linear correlation were used for statistical analysis. RESULTS
Two volunteers were excluded due to noncompliance. The remaining subjects had an age of 51.2 f 5.6 (mean * SD) years, and a body mass of 87.8 kg (SD 6.8). The subjects (n = 5) who were prescribed sucrose overfeeding showed a pre-diet energy intake of 2020 kcal/d (SD 784) (8.5 MJ) per day rising to 2500 kcal (SD 758) (10.5 MJ) according to dietary records. In the other group (n = 5) the basal energy intake was 1650 kcal/d (SD 380) (6.9 MJ) and 2220 kcal/d (SD 580) (9.2 MJ) during the diet period with fat supplementation. In the presentation of the results the subjects are pooled from the following reasons: first, the two groups seemed to be overfed to a similar degree as shown by the diet records and the almost identical increase in body mass. Neither were there any significant differences in sodium excretion. Second, there were no indications of differing responses in catecholamines, thyroid hormones, or plasma insulin to overfeeding with either fat or sugar. Third, in both groups there were similar changes in intraerythrocyte sodium content and ion fluxes (Table 2). During seven days of overfeeding, body mass increased by 0.67 kg (SD 0.6) (Tables 1, 2). The 24-hour urinary sodium excretion did not change significantly between the study periods (Table I). There were no significant changes in urinary output or in
996
FAGERBERG ET AL
Table 2.
Subject (Diet) 1
Individual Chanaes in Bodv Mass, lntraervthrocvte Na
bnmol/L)
Sodium Content
and ion Fluxes
Na-influx fmmol x-’ 1 x h-‘1
Na-effluX rate constant fh-‘)
K ~mmol/Ll
Period
aodv Mass
Total
Control
95.2
10.1
5.8
2.5
0.45
Diet
95.6
9.1
4.4
3.4
0.76
98.6
Control
88.0
11.7
6.1
3.8
0.62
86.4
Diet
88.5
10.2
7.6
2.8
0.37
82.5
Control
81.5
9.2
6.3
3.1
0.50
91.0
Diet
82.5
9.0
6.0
3.4
0.57
96.4
4
Control
80.2
8.1
7.1
2.3
0.35
94.2
(F)
Diet
80.2
8.8
5.0
2.1
0.45
94.9
Control
90.0
9.9
7.1
1.9
0.28
91.8
Diet
91.0
9.5
6.3
2.1
0.35
87.4
Control
81.8
8.9
7.8
1.4
0.19
97.1
Diet
83.4
7.7
4.2
2.7
0.68
93.9
Control
100.3
9.6
5.5
2.5
0.48
90.4
Diet
(CH) 2 (F) 3 (CH)
5 (CH) 6 (F) 7 (CH) 8 (F) 9 (CH)
Exchangeable
100.0
9.9
5.0
3.2
0.65
so.2
Control
86.2
9.9
7.3
2.0
0.29
98.6
Diet
86.8
7.5
5.3
2.7
0.52
92.3
Control
91.0
ND
ND
ND
ND
ND
Diet
92.5
(7.3)
(6.7)
(2.7)
(0.88)
(94.2)
(0.211
(86.6)
10
Control
81.9
(9.1)
(4.9)
(2.81
(F)
Diet
82.3
ND
ND
ND
Mean SD
91.0
ND
ND
Control
87.6
+ 6.6
9.7 ? 0.8
6.6 + 0.8
2.4 k 0.7
0.40
i 0.14
92.1
+ 3.
Diet
88.3
f 6.5t
8.9 k 1.1’
5.5 + 1.1.
2.8
0.54
k 0.15’
92.1
* 4.
k 0.5
Body mass, intraerythrocyte sodium content (total and exchangeable), potassium content, sodium influx, and sodium efflux rate constant before and after overfeeding with 25 energy % fat IF) or carbohydrates (CH). ND - not done. l< 0.05 compared with control values. tc0.01
compared with control values.
plasma concentrations of catecholamines or in serum levels of thyroxine, T3, or reverse T3; nor were there any changes in blood pressure or heart rate (Table 1). Plasma noradrenaline after sustained handgrip and insulin levels showed no significant changes when the diet period was compared to baseline (Table 1). The sodium content (total and exchangeable) of the red blood cells, however, decreased significantly (P < 0.05, respectively) with high energy intake (Table 2). Intraerythrocyte potassium concentration did not change, and no difference in the erythrocyte water content was noted. The sodium efflux rate constant increased significantly after overfeeding compared to the control period (P < 0.05) whereas no significant change in sodium influx was noted. No significant correlation was found between weight gain and increase in efflux rate constant during overfeeding (r = 0.57, P < 0.1). DISCUSSION
In outpatient studies comprising changes in diet it is often difficult for the subjects to adhere strictly to the protocol. Furthermore, it is complicated to evaluate the adherence to the regimen and a general experience is that dietary records often underestimate the energy
intake. This is supported by the results of the present study showing an unexpected low energy intake during basal conditions in both groups. In addition, it has been shown that body mass is a poor indicator of overfeeding with a high interindividual variability probably as a reflexion of dietary thermogenesis.’ In our study, there was no significant change in any parameters during the control period but a significant increase in energy intake according to the dietary records during the diet period. Body mass increased in most patients. In conclusion, we believe that the volunteers were overfed according to the study plan. The present study demonstrates that a moderate increase in energy intake during 1 week is associated with an elevation of the erythrocyte efflux rate constant and a reduction of the red blood cell sodium content. Because we have only measured the total efflux, we are not able to determine whether the increase in efflux rate constant could be ascribed to changes in Na+ KC-pump activity or Na+ K’ cotransport. It has, however, been shown in experiments with different transport inhibitors that the ouabainsensitive Na+ K+ pump covers about 75% of the total sodium efflux while Na’ K+ cotransport only contributes to about 12%~~~It therefore seems reasonable to assume that the present increase in sodium efflux is
SODIUM
997
EFFLUX DURING OVERFEEDING
related to an elevation of the Na+ K+-pump activity but a concomitant influence of the cotransport can of course not be excluded. There are earlier studies on the effects of different caloric intake levels on the Na+ K’ pump. It has been found that five days of sucrose overfeeding causes an increase in Na” K’-ATP-ase mediated ion transport in liver slices and intact soleus muscles in normal mice.14 Using an oubain-inhibitable ATP-ase assay on human erythrocyte vesicles Mir et al found an increase in sodium-potassium ATP-ase activity 60 minutes after a meal with a high interindividual variability.26 DeLuise and co-workers on the other hand, found no effects on the number or activity of Na’ K’-ATP-ase units in erythrocytes measured by (3 H) oubain binding and ?ubidium uptake after acute feeding on a hypocaloric diet, fasting, and refeeding in humans.*’ The discrepancy between these studies illustrates not only the problem of differing methods but also the question of whether the erythrocyte is a representative model for transmembrane ion fluxes in all tissues. When comparing studies in the Na’ K’ ATP-ase activity in liver tissue on one hand and in red blood cells on the other, conflicting results have been reported concerning human obesity28.29as well as hyperthyroidism.30,3’ ,4gainst the background that the sodium-potassium pump activity represents most of the total sodium efflux and that the pump in different tissues (muscle, liver) has been shown to be influenced by insulin,” thyroid hormone,6 and catecholamines” we found it interesting to study the dietary effects on these hor-
mones especially because overfeeding is associated with an increase in circulating insulin’ and T32 with decrease in reverse T3.* The present study, however, showed no such changes concerning plasma insulin, T3, T4, or reverse T3. Young and Landsberg have reported that noradrenaline turnover rises in response to overfeeding in rats.“’ Recent short-term studies on humans have shown that overfeeding increased noradrenaline turnover,4 peak standing plasma noradrenaline, and resting metabolic rate.3 We, however, found no evidence of an increase in sympathetic nervous system activity because there were no significant changes in either urinary noradrenaline output or plasma noradrenaline during supine rest or isometric exercise. Nor were there any changes in heart rate, systolic or diastolic blood pressure, as were found in the studies mentioned above.3,4 The urinary adrenaline excretion was not affected by increasing the energy intake. The fact that we could not detect any differences in insulin and thyroid hormone levels or in indices of sympathetic nervous system activity does not exclude the possibility that there may be such a differing response to overfeeding with sugar or fat because our study groups were small. In conclusion, we have found that moderate overfeeding induces an increased sodium efflux and a decreased sodium content in erythrocytes from slightly obese men. Because no estimation of expenditure was obtained our findings can not be related to thermogenesis.
REFERENCES I Danforth Jr E, Burger AG, Goldman RF, et al: Thermogenesis during weight gain, in Bray G (ed): Recent Advances in Obesity Research: II. Proceedings of the 2:nd International Congress on Obesity. London, Newman Publishing Ltd, 1978, pp 229-236 2 Danforth Jr E, Horton ES, O’Connell M, et al: Dietary induced alterations in thyroid metabolism during overnutrition. J Clin Invest 64:1336-1347, 1979 3. Welle S, Campbell RG: Stimulation of thermogenesis by carbohydrate overfeeding. Evidence against sympathetic nervous system mediation. J Clin Invest 7 I:9 16-925, 1983 4. O’Dea K, Esler M, Leonard P, et al: Noradrenaline turnover during under- and over-eating in normal weight subjects. Metabolism 3 I :896-899, I982 5. Sims EAH, Danforth Jr E, Horton ES, et al: Endocrine and metabolic effects of experimental obesity in man. Ret Prog Horm Res 291457496, 1973 6. Smith TJ, Edelman IS: The role of sodium transport in thyroid thermogenesis. Fed Proc 38:215&2153, 1979 7. Ismail-Beigi F, Edelman IS: The mechanism of the calorigenic action of thyroid hormone. Stimulation of Na’, K+-activated adenosine-triphosphatase activity. J Gen Physiol 57:710-722, 1971 8. Whittam R: Active cation transport as a pace-maker of respiration. Nature 191:603-604, 1961 9. Himms-Hagen J: Cellular thermogenesis. Ann Rev Physiol 38:315-351, 1976
10. Clausen T, Flatman J: The effect of catecholamines transport and membrane potential in rat soleus muscle. 270:383-414, 1977
on Na-K J Physiol
11, Clausen T. Kohn PG: The effect of insulin on the transport of sodium and potassium in rat soleus muscle. J Physiol 265:19--42. 1977 12. Welle S, Lilavivat U, Campbell RG: Thermic effect of feeding in man: Increased plasma norepinephrine levels following glucose but not protein or fat consumption. Metabolism 30:953-968, 1981 13. Landsberg L, Young JB: Diet and the sympathetic nervous system: relationship to hypertension. Int J Obesity 5 (suppl) 1:7991. 1981 14. Flier JS, Usher P, DeLuise M: Effect of sucrose overfeeding on Na, K-ATP-ase-mediated 86Rb uptake in normal and ob/ob mice. Diabetes 30:975-978, 1981 15. Lindberg W, Natwig H, Rygh A, et al: Hoyde-og vektundersogelser hos voksne menn og kvinner. Forslag till nye norske hoydevektnormer. Tidskr Norske Laegefor 76:361--368, 1956 16. Blomqvist N, Lindstedt G, Lundberg P-A, et al: Noinhibition by Li+ of thyroxine monodeiodination to 3,5,3’-triiodothyronine and 3,3’.5’-triiodothyronine. Clin Chim Acta 79:4577464. 1977 17. Hjemdahl P, Daleskog M, Kahan T: Determination of plasma catecholamines by high performance liquid chromatography with
998
electrochemical detection: Comparison with a radioenzymatic method. LifeSci 25131-138, 1979 18. v Euler US, Lishajko F: The estimation of catecholamines in urine. Acta Physiol Stand 45:122-131, 1959 19. And& L, Piros S, Hansson L, et al: Different haemodynamic reaction patterns during noise exposure in normotensive subjects with and without heredity for essential hypertension. Clin Sci 63:371-374, 1982 20. Herlitz H, Jonsson 0, Elmfeldt D, et al: Intraerythrocyte sodium content and transmembrane fluxes in patients treated with diuretics or beta-blockers. Acta Pharmacol Toxic01 54 (Suppl) 1:3741,1984 21. Henningsen NC, Nelson D: Net influx and efflux of 22Na in erythrocytes from normotensive offspring of patients with essential hypertension. Acta Med Stand 210:85-91, 1981 22. Jonsson 0, Herlitz H, Gudmundsson 0, et al: Fluxes of sodium and potassium in human erythrocytes: Determination based on computer analyses of isotope uptake curves. Acta Physiol Stand 121:93-95,1984 23. Naucler J, Jonsson 0, Lundstam S, et al: Transmembrane fluxes of potassium in dog kidney slices. A quantitation of the effect of warm ischemia. Eur Surg Res 9:384-396, 1977 24. Mir MA, Bobinski H: Altered membrane sodium transport
FAGERBERG ET AL
and the presence of a plasma oubain-like inhibitory factor in acute myeloid leucemia. Clin Sci Mol Med 48:213-218, 1975 25. Walter U, Distler A: Abnormal sodium efflux in erythrocytes of patients with essential hypertension. Hypertension 4:205-210, 1982 26. Mir MA, Charalombous BM, Morgan K: Erythrocyte sodiurn-potassium-ATP-ase and obesity. N Engl J Med 306:809-810, 1982 27. DeLuise M, Izumo H, Grace EE, et al: Effect of diet upon the erythrocyte Na, K pump. J Endocrinol Metab 56:739-743, 1983 28. Bray GA, Kral JG, Bjorntorp P: Hepatic sodium-potassium dependent ATP-ase in obesity. N Engl J Med 304:1580-l 582, 198 1 29. De Luise M, Blackburn GL, Flier JS: Reduced activity of the red cell sodium-potassium pump in human obesity. N Engl J Med 303:1017-22, 1980 30. Gambert SR, Ingbar SH, Hagen TC: Interaction of age and thyroid hormone status on Nat K+ ATP-ase in rat renal cortex and liver. Endocrinology 108:27-30, 1981 31. Cole CH, Waddel RW: Alteration in intracellular sodium concentration and oubain-sensitive ATP-ase in erythrocytes from hyperthyroid patients. J Clin Endocrinol Metab 42:1056-1063, 1976